module i. introduction to quality management lecture – 1 ... · another quality guru, j. m....
TRANSCRIPT
Module I. Introduction to Quality Management
Lecture – 1 How the concept of Quality Management evolved over time?
From ancient time, quality of goods and services are monitored directly or indirectly. If we
look at construction of the pyramid, Greek ancient arts, crafts, and architectures, Roman-built
cities, it clearly demonstrates artists and engineers commitment for achieving the excellence
in quality. However, till 1800, production of goods and services was primarily done by small
group of individuals. These small groups were often family businesses. Thus, the standard of
quality was controlled and set by individual who was in turn also responsible for producing the
item. This phase, comprising the time period up to 1900, is called the period of ‘Operator Quality
Control’. The entire product was manufactured by a single person (or operator) or by small group
of persons, who essentially controlled quality. Thus, controlling and improving quality of the
product was aligned with the philosophy of pride in workmanship.
From early 1900s to 1920, a second phase evolved, which called the ‘Foreman Quality Control’
period. In this phase, the concept of mass production with little emphasis on personal
accomplishment at work place was introduced. Supervisors are responsible to ensuring that
quality was achieved. Foremen or supervisors controlled the quality of the product, and they
were also responsible for the shop floor operations.
The period of 1920 to 1940 saw the next phase of quality. This phase was so-called ‘Inspection
Quality Control’. With more complicated products and processes it became impossible for to
keep close watch over individual stages of operation. Inspectors were assigned to check the
quality of a product after processing. Individual product standards were set and any
discrepancies between standard and actual product features was reported. Defective items were
set aside as scrap, and few items with minor defects are reworked to meet the specified standard or
specification. In this period, statistical process control aspects of quality were also popularized,
and gained widespread application in industries. In 1924, Walter A. Shewhart of Bell Telephone
Laboratories introduced the concept of statistical charts to monitor variability of product
characteristics. These charts were called control charts. In the latter half of 1920s, H. F. Dodge
and H. G. Romig, also from Bell Telephone Laboratories, proposed acceptance sampling plans
for inspection. These plans substituted the concept of 100 percent inspection. During 1930’s
application of acceptance sampling plans was in full flow in industries. In 1929, Walter
Shewhartwith the help of American Society for Testing Materials (ASTM), American Society of
Mechanical Engineers (ASME,), American Statistical Association (ASA), and Institute of
Mathematical Statistics (IMS) created the joint committee for the development of statistical
techniques for application in engineering industries.
The phase of ‘Statistical Quality Control’ was between 1940 and 1960.During World War II, the
principles of sampling inspection plan was extremely useful. The American Society for Quality
Control (ASQC) was formed in 1946. A set of sampling inspection plan for attributes, so-called
MIL-STD-105A was developed in 1950. These plans underwent various modifications, viz.
MIL-STD-105B, MILSTD-105C, MIL—STD-105D, and MIL-STD-105E. In addition, during
1957, a set of sampling plans for variables called MIL-STD-414 was also proposed. Juran
published his Quality Control Handbook in 1957.
Use of quality control procedures and benefits of statistical quality control was not explored in
most of the U.S. industries. This may be due to monopoly market. However, Japan after World
War II, embraced the new philosophy wholeheartedly. Edwards Deming was invited to Japan
during 1950, and Japanese engineers were convinced about the importance of statistical quality
control as a means to gaining competitive advantage in world economy. Another quality guru, J.
M. Juran, visited Japan in 1954 and further impressed upon the strategic role that management
plays to achieve end quality. Thus, they started to develop strong commitment to train and educate
their employees on statistical process control.
The next phase of quality during 1960 is known as Total Quality Control. An important feature
during this phase was involvement of several departments and personnel in the quality
development process. Prior to this period, the attitude was quality is the responsibility of the
inspection. In 1960s, there was a change in this attitude. Employees began to understand that
each department within an organization has a contribution to build quality in an item. Concept
of zero defects, which encircle around achieving productivity through worker involvement,
emerged during this period. With more or less same underlying philosophy, quality circles were
introduced in many Japanese industries. The concept of quality circles is based on participative
or team work style of management. It believes that quality and productivity can be achieved
through informal group discussion, decision, and pertinent action.
1970 is the phase of ‘Total Quality Management’. This phase involved the participation of
everyone in the organization, from the operator to supervisor, manager, and even the chief
executive officer. Quality was responsibility of every individual. Feigenbaum, another quality
guru, defines the philosophy as:
‘A quality practice that is agreed on companywide and plant wide operating work structure,
documented in effective, integrated technical and managerial procedures, for guiding the
coordinated actions of the people, the machines, and the information of the organization in the best
and most practical ways to assure customer quality satisfaction and economical costs of quality.’
1970 also marked the extensive use of a graphical tool known as the cause-and-effect diagram.
Also in this decade, G. Taguchi of Japan introduced the concept of robust design in statistical
experimentation.
During 1980s, various quality control and statistical software came into the market. The notion
of a total quality management increased the emphasis on supplier’s quality, product design,
quality assurance. Ford Motor, Daimler Chrysler and General Motors Corporation adopted the
quality philosophy and insisted supplier to adopt various quality control and quality improvement
techniques.
In 1989, Motorola started the Six Sigma initiative, a quality philosophy driven by statistical
approach for decision making, which within 10 years was sincerely adopted by various other
companies.
Module I. Introduction to Quality Management
Lecture 2 - What do we mean by Product and Service Quality?
‘Quality’ can be defined as a standard measure of how well a product or service conforms to the
specified standards, so as to meet the customer requirements. Quality has been defined by
various quality gurus. Juran defined quality as ‘fitness for use’ in 1974. Crosby, in 1979, defined
the quality as ‘conformance to requirements or specifications.’ Garvin, in 1984 divides the
definition of quality into five major categories—namely, transcendent, product-based, user-
based, manufacturing-based and value-based. In addition, he also identifies eight attributes
that may be used to define product quality: performance, features, reliability, conformance,
durability, serviceability, aesthetics, and perceived quality. The definition proposed by Crosby
seems more appropriate from both service and manufacturing perspective. However, terms such
as delighting customers, robustness, reducing variability can also be associated when
organization talk about quality. The driving force to determine the level of quality that should
be designed into a product of service is ‘The customer’. Quality also has a time dimension.
In other words, as the need and preference of customer changes with time, the level of quality
or degree of customer satisfaction also changes. Thus, quality, in this sense, is not constant.It is
a crucial parameter that differentiates an organisation from its competitors.Thus, term quality
also implies different levels of expectations from different segment of consumers.
Now, who are the customers?
There are two distinct types of customers- external and internal. An external customer may be
the one who uses the end product or service, the one who purchases the end product or service,
or the one who influences the sale of the end product or service. An external customer exists
outside the organization.
An internal customer is also important. Every function within an organization, whether it is
engineering, order processing, or production has an internal customer. That means each function
receives a product or service from another function and, in exchange, provides a product or
service to a subsequent function. From process perspective, every process is considered a
customer of the preceding process. Grinding process can be internal customer to boring process.
A diagrammatic representation is provided below (Figure1-1) to show internal and external
customer in the supply network perspective.
Figure 1-1: Customer from basic Supply network Perspective
From the customers' perspective, service is the combination of the customers experience and
their perception of the outcome of the service. The term service quality (Johnston and Clark,
2008) is often used to mean different aspect. Some may use the term to mean how the customer
is treated overall. This may be more accurately called ‘quality of service’, as opposed to service
quality, which can mean the entirety of outcome and experience. Sometimes service quality is
used to mean the same as satisfaction, i.e. perceived service quality.
Few definitions, which are important while we discuss on product quality is worth
mentioning:
Quality Characteristics
There can be more than one element that defines the intended quality level of a
product or service. These elements are so-called ‘quality characteristics’. Quality
characteristics may be of several types (Montgomery and Runger, 2010). It may be
Physical: length. weight, voltage, viscosity, or
Sensory: taste, appearance, color, or
Time Orientation: reliability, durability, serviceability
Thus, the gross weight of a coke cans, the tensile strength of a bar, the specific
gravity of a liquid, and so on. There may be some intangible characteristics, such as the
taste, smell beauty. There can also be ethical characteristics such as honesty, courtesy,
friendliness which are difficult to measure and define.
Variables and Attributes
Quality characteristics can fall into two broad classes, viz. variables and attributes.
Characteristics that are measurable and are expressed on a numerical scale (ordinal or interval)
are called variables. The diameter of a bore expressed in millimeters is a variable, or density of
a liquid in grams per cubic centimeter or customer satisfaction expressed in a scale of 7. If we
express characteristics only in terms of conforming and defective, it falls in the attribute or
nominal category.
Defects and Defective Unit
Before defining an attribute, the terms defect and a defective unit should be defined.
Defect is a quality characteristic that do not meet its stipulated specifications. Let's say the
specifications of the thickness of steel washers are 3 ± 0.1 millimeters (mm). If we have a
washer with a thickness of 3.15 min, then its thickness is a defect.
The American National Standards Institute (ANSI) and the American Society for Quality Control
(ASQC) provides a definition of a defect as stated in ANSI/ASQC Standard:
‘A defect is a departure of a quality characteristic from its intended level or state that occurs
with a severity sufficient to cause an associated product or service not to satisfy intended
normal or reasonably foreseeable usage requirements.’
A defective unit has one or more defects such that the unit is unable to meet the intended
standard or use and is unable to function as required. An example of a defective unit might be a
cast iron cylinder that has an internal diameter and a weight that both fail to satisfy
specifications, thereby making the unit dysfunctional.
A quality characteristic is said to be an attribute if it is classified as either conforming or
defective to a stipulated specification. A quality characteristic that is not measured on a
numerical scale is expressed as an attribute. For example, the smell of a perfume is characterized
as either acceptable or not; the color of a cloth is either acceptable or not. Variables are treated
as attributes because of their simplicity to measure them this way or because it is difficult to
obtain data on them. Many examples may be cited in this category. For instance, the diameter of a
engine cylinder is, in theory, a variable. However, if we measure the diameter using a go/no-go
gauge and classify it as either conforming or defective (with respect to some established
specifications), then the characteristic is expressed as an attribute. The reasons for using a
go/no-go gauge, as opposed to a micrometer, could be economical. A measurement by go/no-
go gauge may be much shorter and consequently less expensive.
Standard or Specification
As the definition of quality involves meeting the requirements of the customer, these requirements
need to be documented. A standard, or a specification, refers to a precise statement that validates the
requirements of the customer; it may relate to a product, a process, or a service. For example, the
specifications for a bore might be 3 ± 0.1 centimeters (cm) for the inside diameter, 5 ± 0.2 cm for the
outside diameter, and 12 ± 0.5 cm for the length. This means that for the bore to be acceptable to the
customer, all the three dimensions must be within the specified limit.
Module I. Introduction to Quality Management
Lecture 3 - What are the dimensions of quality?
Before we discuss on dimensions of quality, we must discuss three aspects associated with
definition of quality: quality of design, quality of conformance, and quality of performance.
Quality of Design
Quality of design is all about set conditions that the product or service must minimally have to satisfy
the requirements of the customer. Thus, the product or service must be designed in such a way so
as to meet at least minimally the needs of the consumer. However, the design must be simple and
also less expensive so as to meet the customers' product or service expectations. Quality of design is
influenced by many factors, such as product type, cost, profit policy, demand of the product, avail-
ability of parts and materials, and product reliability.
Quality of Conformance
Quality of conformance is basically meeting the standards defined in the design phase after the
product is manufactured or while the service is delivered. This phase is also concerned about quality
is control starting from raw material to the finished product. Three broad aspects are covered in this
definition, viz. defect detection, defect root cause analysis, and defect prevention. Defect prevention
deals with the means to deter the occurrence of defects and is usually achieved using statistical process
control techniques. Detecting defects may be by inspection, testing or statistical data analysis collected
from process. Subsequently, the root causes behind the presence of defects are investigated, and finally
corrective actions are taken to prevent recurrence of the defect.
Quality of Performance
Quality of performance is how well the product functions or service performs when put to use. It
measures the degree to which the product or Service satisfies the customer from the perspective of
both quality of design and the quality of conformance. Meeting customer expectation is the focus
when we talk about quality of performance. Automobile industry conduct test drive of vehicles to
collect information about mileage, oil consumption. Bulbs are life tested to understand its reliability
during useful life. Customer survey is conducted to find customer’s perception about service
delivered. If product or service does not live up to customer expectation, then adjustments are needed
in the design or conformance phase.
Garvin (1984) also provides discussion of eight critical dimensions of product quality. The
summarized key points concerning these dimensions of quality is provided below.
Performance (will the product do the intended job in field?)
This we have already discussed. It talks about evaluation of product or service performance with
respect to certain specific functions and determine how well it performs from customer’s
perspective.
Reliability (how often the product can fail within a stipulated time?)
It talks about probability of not failing of components of say automobiles or airbus while on
service for a specified time period. Less the reliability, more the chances of repair or
replacement.
Durability (how long can the product last?)
This is the effective life of the product or longevity before it is declared as unfit for use. Repair is
not possible after this phase of life.
Serviceability (how easy is it to repair the product?)
Customer's view on quality is also influenced by how quickly and economically a repair or
routine maintenance activity can be accomplished. This is mentioned as serviceability. For
examples how long did it take to correct an error in your credit card statement by the bank?
Aesthetics (how appealing does the product look like?)
This is all about visual appeal of the product, often taking into account factors, such as style,
color, shape, packaging, tactile characteristics, and other sensory features.
Features (value or what does the product can actually do?)
Customers tend to purchase products that have more value added features. This can be beyond
basic criteria to enter into the market. A spreadsheet package may come with built-in statistical
quality control features while its competitors did not in the same price range. Feature may also be
definite as addition or secondary characteristics attached and supplements primary functionary of
a product. Thus, car stereo is a feature of an automobile whose primary function is transportation.
Perceived Quality (what is the customer’s feeling about the product after intended use?)
This is all about impression of a customer after using the product and/or service. This dimension
is directly influenced by any failures of the product that are highly visible to the public or the
way customer is treated when a quality-related problem with a product is addressed. Customer
loyalty and repeated business are closely related with perceived quality. For example, if you
make regular business trips by a particular airline, which almost always arrives late with few
incidence of luggage lost in transit, you will probably prefer not fly on that carrier and prefer its
competitor. So you will rate this dimension very low for such carrier.
Conformance to Standards (is the product made exactly as the designed
?)
This is what was discussed earlier as quality of conformance.
Service Quality
Service is generally defined as an experience felt by the consumer. Say, in a restaurant, the way
the customer is treated is considered as a service. Services are often intangible in nature. The
quality of service is judged by how well the customer is satisfied with the service. Service
quality is about comparing performance with the customer expectations. Service quality also
leads to customer satisfaction and interrelated. The key to retain customers is to understand their
needs and fulfill those needs. Making customers buy the services repeatedly requires focus on
dimensions of service quality. There are five dimensions of service quality and given below:
Tangibles: The tangible dimension of quality is related to the surroundings in which the service
is provided to the customers. In a restaurant, it may be seating arrangement, interior decoration
and lighting arrangement.
Reliability: Reliability refers to the dependability of customers on specific service. It is all about
what is promised and what is delivered. Like, Indigo airlines in India have proved to be low cost
airlines with high punctuality.
Responsiveness:Responsiveness refers to the time taken by a service provider to respond to
request. Like, LG customer care in India promises response to customer complaints within 24
hours.
Assurance: This dimension of service quality is related to the competence of the service
employee. The employees must be competent to gain the trust of customers.
Empathy: Empathy refers to caring attitude that an organization shows toward customer. This
dimension of service quality calls for individual attention to customer, so as to make them feel
special.
Considering the above dimensions, comparisons are made between actual service performance
and expectations of customers. The difference between customers’ expectations and actual
delivery (so-called ‘perception’) at the time of service performance is known as service quality
gap.Organization conduct survey and exploratory research to study the various service gaps, so
as to understand why the gap arises and how it can be reduced. Readers may refer Parasuraman
et al.(1985, 1988) paper for further details on Gap Models.
Module I. Introduction to Quality Management
Lecture 4 - What are the underlying quality philosophies suggested by Deming, Juran and
Crosby?
Several quality guru’s, such as W. Edwards Deming. Philip B. Crosby, and Joseph M. Juran, has
made significant contributions in the field of quality. They are largely responsible for the global
adoption and integration of quality management in industry. They preached that management
commitment is the key to a successful program in quality. They also emphasized that any philosophy
to improve quality in a company needs time and cannot occur overnight.
W. Edwards Deming is credited with the impressive turnaround in Japanese industry after World War
II. Deming's philosophy emphasizes the role of ‘management of the problems’ that industry
faces. Deming said that about 85% of the problem can be solved only by management. These
involve changing the method of operation and are not by scolding the workers. His idea was
to improve process and not on blaming or scolding workers. In Deming's world, workers'
responsibility lies in communicating to management the information they possess regarding
the process and both must work in harmony. The Deming's ideal management style is holistic
and organization is viewed as an integrated entity. The idea is to plan for the long run and
provide a course of action for the short run. Deming believed in the adoption of a total
quality philosophy and emphasized the never ending nature of statistical quality control in
the quality improvement process. Deming's approach demands a cultural transformation in
the organization. Deming advocated certain key components that are essential for the
journey toward continuous improvement, viz. Knowledge of the system and the theory of
optimization(to look into system as a whole and not as individual process), Knowledge of the
theory of variation( understanding common and special cause of variation and emphasize on
statistical process control), Exposure to the theory of knowledge(data driven prediction that
is based on underlying knowledge about processes), Knowledge of psychology (understand
the behavior and interactions of people and also the interactions of people with their
working environment). Deming provides 14 points for management that will sustain
productivity and competitiveness of the company in the long run. Book by Deming (published in
1982, 2000)or any book on Quality Management can be seen by readers to understand in-depth the
14 points. In around 1950, Shewart cycle was renamed in Japan as Deming PDCA cycle. This is a
continuous cycle of process improvement. This is illustrated in Figure 1-2 given below.
Figure 1-2 PDCA Deming Cycle
Deming's 14 points for management provide a road map for continuous quality improvement.
While implementing these points, certain practices of management are
labeled by Deming as deadly diseases or sins. These are (i) management by visible figures only,
(ii) lack of constancy of purpose, (iii) performance appraisal by numbers, (iv) a short-term view
of organization, and (v) mobility of management. These must be eliminated. Most of Deming's
deadly diseases involve a lack of understanding of variation.
Joseph. M. Juran emphasized on seven step process for controlling quality, which is employed
by various organizations to control the processes.In this context, Juran first visited Japan in
1950’s,and educated the management of large organistions about the need of management’s
commitment to attain quality. The quality standards developed by Japanese are based on these
concepts. According to Juran philosophy,Quality is defined as “fitness for use”.
Philip B. Crosby has a particularly wide-ranging understanding of the various operations in
industry because he started as a line inspector and worked his way up. Such firsthand experience
has provided him with a keen awareness of what quality is, what the obstacles to quality are, and
what can be done to overcome them. He founded Philip Crosby Associates in 1979. His quality
management grid identifies and pinpoints operations that have potential for improvement. The
grid is divided into five stages of maturity, and six measurement categories aid in the evaluation
process. Readers can refer to his book ‘Quality is Free (1979)’ for in-depth about his philosophy.
He suggested that the rational quality improvement approach is to prevent defects. He defined
that the only performance standard is zero defect. Crosby emphasized on performance from the
cost of quality perspective. He preached to reduce costs of unquality, such as scrap, rework,
inventory, machine breakdown, inspection, etc. These are the cost that leads to poor quality.
Module I. Introduction to Quality Management
Lecture 5 -What do we mean by Quality Cost?
Quality costs are defined as those costs that are associated with the non-achievement of product
or service quality as defined by the requirements established by the organization and its contracts
(agreements) with customers. In simple terms, quality cost is the cost incurred by the firm
because of producing poor quality products. Measurement and analysis of various cost aids in
tracking the impact of an effective quality management system. Quality costs can be summed up
as costs of preventing of non-conformance of requirements, inspecting product/service for non-
conformances and failure in meeting specifications. The American Society for Quality Control
(1971) has defined four major categories for quality costs, which are provided below:
Prevention Costs
Prevention costs are incurred in planning, implementing, and maintaining of a quality practice. It
include salaries and developmental costs for process control approaches, information systems,
and all other costs associated with making the product right the first time. Also, costs associated
with education and training is included in this category. Defect identification and removal and
the cost of a quality audit are included in the prevention cost.
External Failure Costs
External failure costs are incurred when the product does not perform satisfactorily
after it is shipped to the end customer. If there are no defective units, the external failure cost can
be zero. However, cost incurred due to customer complaints, costs of investigation and
adjustments if required, and those associated with receipt, handling, repair (if possible), and
replacement of defective products comes within the external failure cost. Warranty cost (failure
of a product within the warranty time) which is specifically monitored in industries also fall
under this category.
Appraisal Costs
Appraisal costs are related with measuring, evaluating, or inspecting products, components, or
purchased materials to determine their degree of conformance to specified design standards.
Such costs include dealing with the inspection and test of incoming materials as well as product
inspection and testing at various stages of manufacturing till final acceptance. Appraisal costs are
associated with managing the outcome, whereas prevention costs are associated with managing
the goal.
Internal Failure Costs
Internal failure costs are incurred when products, sub assemblies, components or materials fail to
meet quality requirements prior to the transfer of ownership to the internal customer. These costs
will disappear if there were no defects or defective in the product while it is manufactured in-
house. Internal failure costs also include labor and overhead cost associated with any internal
repair.
Module II. Process Quality Improvement
Lecture -1 Why process quality improvement is important?
From operations management perspective, a process is any activity or group of activities that
takes one or more inputs, transforms them, and provides one or more outputs for its customers
(internal or external). The key to success in an organization is to understand how their processes
work to deliver the required outputs. Any process should add value and unnecessary waste
activities should be eliminated from the process steps, as per definition in Lean Management
philosophy. In the context of Quality Management philosophy, process is transformation of
inputs into output, which satisfies the required Quality Characteristics defined by the customers.
These characteristics are called ‘CTQ’s’ (Critical-to-Quality) or ‘responses’. The transformation
happens by controlling few vitals critical input and process variables (x1...xp) known as
controllable variables. These variables actually influence the mean and variance of the CTQ.
Thus proper setting of these variables is critical to get the best or optimal output.
However, there are other variables (z1…zm) which cannot be controlled, say room temperature,
humidity, or uneconomical to control. The variation caused in CTQ by these variables is
assumed to be the natural variability or chance cause variability. Taguchi emphasized to
minimize CTQ variability even in presence of these uncontrollable or noise variables by using
orthogonal array design based DOE.
Figure 2-1 Schematic diagram of a Process with Influential Variables
Uncontrollable Variables
z1 z2 zm
x1 x2 xp
CTQ (s) or
Responses (y)
Inputs
Controllable Variables
Determining the best setting for controllable variables is the primary focus of process quality
improvement activity. If process improves, we will get the best output or responses and as a
consequence best desired quality product for the end customer. Every company focuses on
process quality improvement so as to improve their prime competitive priority (or Quality).
Module – 2 Process Quality Improvement
What are the graphical and statistical techniques commonly used for understanding
current state of quality? What are the process quality monitoring, control and
improvement techniques?
Systematic solution approach to any quality improvement activity is critical and always
emphasized by quality gurus (Juran, Deming, and Shewart). Various tools and techniques are
commonly used to identify the critical control variables. The very basic techniques used in
quality management is 7 QC Tools, which consist of Pareto Diagram, Process Flow
Diagram, Cause and Effect Diagram, Check Sheets, Histogram, Run Charts, and Scatter
diagram. Additional statistical tools used are hypothesis testing, regression analysis,
ANOVA (Analysis of Variance), and Design of Experiment (DOE). In the following
section, we will go through each and every technique in a greater detail.
7QC TOOLS
Pareto Diagram
Alfredo Pareto (1848-1923) conducted extensive studies of distribution of wealth in Europe. He
found that there were a few people with a lot of money and majority of the people are having
little money in their hand. This unequal distribution of wealth became an integral part of
economic theory. Dr. Joseph Juran recognized this concept as a universal concept which can be
applied to many other fields. He coined the phrase ‘vital few and useful many’.
A Pareto diagram is a graph that ranks data (on say types of defects) in descending order from
left to right, as shown in Figure 2-2. In the diagram, data is classified as types of coating
machines. Other possible data classifications include problem, complaints, causes,
nonconformities types, and so forth. The vital few will come on the left of the diagram, and
useful many are on the right. It is sometimes worthy to combine some of the useful many into
one classification called "other". When this category is used, it is placed on the far right.
The vertical scale can be dollar value (or frequency), and percentage in each category is shown
on top of each bar. In this case, Pareto diagrams were constructed for both frequency and dollar
value. As can be seen from the figure, machine 35 has the greatest number of nonconformities,
but machine 51 has the greatest dollar value. Pareto diagrams can be distinguished from
histograms (to be discussed) by the fact that horizontal scale of a Pareto diagram is categorical,
whereas the scale for histogram is numerical or continuous.
Figure2-2 Simple Pareto Diagram
Pareto diagrams are used to identify the most important problem type. Usually, 75% of the
problems are caused by 25% of the items. This fact is shown in the above figure, where coating
machines 35 and 51 account for about 75% of the total non-conformities.
Actually, most important items could be identified by listing them in descending order. However,
graph has an advantage of providing a visual impact, showing those vital few characteristics that
need attention. Construction of a Pareto diagram is very simple. There are five steps involved:
Step-1: Determine method of classifying data: by problem, cause, nonconformity, and so forth.
Step-2: Decide if dollars (best), frequency, or both are to be used to rank the characteristics.
Step-3: Collect data for an appropriate time interval or use historical data.
Step-4: Summarize data and rank order categories from largest to smallest.
Step-5: Construct the diagram and find the vital few problem area.
The Pareto diagram is a powerful quality improvement tool to determine the most critical
problem to be considered first. The diagram can also provide cumulative % information and
given in many statistical software (say, MINITAB, http://www.minitab.com/en-
us/products/minitab/?WT.srch=1&WT.mc_id=SE001570) as shown below.
Frequency 74 57 52 34 33 23 11Percent 26.1 20.1 18.3 12.0 11.6 8.1 3.9Cum % 26.1 46.1 64.4 76.4 88.0 96.1 100.0
Defect Type OtherDentsDentPoor SealScratchFinishO-Ring
300
250
200
150
100
50
0
100
80
60
40
20
0
Freq
uenc
y
Perc
ent
Pareto Showing Cumulative Percentage of Defects
Figure2-3 Pareto Diagram with Cumulative %
Process Flow Diagram
For many products and services, it may be useful to construct a process flow diagram. Figure 2-
4 shows a simple process flow diagram for order entry activity of a make-to-order company that
manufactures gasoline filling station hose nozzles. These diagrams show flow of product or
service as it moves through various processing stages. The diagram makes it easy to visualize the
entire multistage process, identify potential trouble spots, waste activities, and locate control
points. It answers the question, "Who is our next customer?" Improvements can be accomplished
by changing (reengineering), reducing, combining, or eliminating process steps.
Telephone
Log in
Letter
Fax CreditCheck
ContactReview
Hold
InventoryCheck
ScheduleProduction Production
Notify Drawing
Figure 2-4 Process Flow diagram for an order entry activity
Standardized symbols
(http://www4.uwsp.edu/geo/faculty/gmartin/geog476/Lecture/flowchart_symbols.html) may be
used as recommended by industrial engineering and Lean Management (Value Stream Mapping,
http://www.strategosinc.com/vsm_symbols.htm) text book. In Six Sigma methodology, process
mapping is done by SIPOC (Suppliers-Inputs-Process-Outputs-Customer,
http://www.isixsigma.com/tools-templates/sipoc-copis/sipoc-diagram).
Cause and Effect Diagram
A cause-and-effect (C&E) diagram is a picture composed of lines and symbols designed to
represent a meaningful relationship between an effect (say Y) and its potential causes (say X).
Potential causes (which have evidence) are not all possible causes that come up in brain storming
exercise. It was developed by Dr. Kaoru Ishikawa in 1968, and sometimes referred to as the
‘Ishikawa diagram’ or a ‘fish bone diagram’.
C&E diagram is used to investigate either a "bad" effect and to take action to rectify the potential
causes or a "good" effect and to learn those potential causes that are responsible for the effect.
For every effect, there are likely to be numerous potential causes. Figure 2-5 illustrates a simple
C&E diagram with effect on right and causes on left. Effect is the quality characteristic that
needs improvement. Causes are sometimes broken down into major sub causes related to work
method, material, measurement, man (people), machinery (equipment), and environment (5M &
1E). It is not necessary that every diagram will always have 5M and 1 E cause and can depends
also on the problem type. There can be other major causes in case of service-type problem.
Each major cause is further subdivided into numerous sub causes. For example, under work
methods, we might have training, knowledge, ability, physical characteristics, and so forth. C&E
diagrams are the means of picturing all these major and sub causes. The identified potential
causes considered critical (say 1, 2, 3, 4 and 5 as given in the below diagram) may be further
explored by experimentation to understand their impact on the house paint.
Figure 2-5: A Simple Cause and Effect Diagram
C & E diagrams are useful to
1) Identify potential causes and not all possible causes,
2) Analyze actual conditions for the purpose of product or service quality improvement
3) Eliminate conditions which cause nonconformities and customer complaints.
4) Statistical Experimentation, Decision-making and corrective-action activities.
C& E diagram can also be generated in MINITAB as shown below.
FlawsSurface
Environment
Measurements
Methods
Material
Machines
Personnel
Testing
Mentors
O perators
Training (Inhouse)
Superv isors
Shifts
C ondition
A ccuracy
Speed
Lathes
Bits
Sockets
Suppliers
Lubricants
A lloy s
Erratic
Too slow
Brake
Engager
A ngle
Inspection procedure
Microscopes
Micrometers
Moisture C ontent (%)
C &E Diagram in MINITAB
Figure 2-6: Cause & Effect Diagram in MINITAB
Check Sheets
Main purpose of check sheets in earlier days is to ensure that data was collected carefully and
accurately by concerned personnel. Data is to be collected in such a manner that it can be quickly
and easily used and analyzed. The form of check sheet is individualized for each situation and is
designed by the project team. Figure 2-7 shows a check sheet for paint nonconformities for
bicycles.
Check sheets can also be designed to show location of defects. For example,
check sheet for bicycle paint non conformities could show an outline of a bicycle,
with ‘X’s indicating location of nonconformities. Creativity plays a major role in
design of a check sheet. It should be user-friendly and, whenever possible, include
information on location.
Figure 2-7: A Typical Check Sheet
Histogram
Histogram provides variation information of the characteristic of interest, as illustrated by
Figure 2-8. It suggests probability distribution shape of the sample observation and also
indicates possible gap in the data. Horizontal axis in Figure 2-8 indicates scale of measurement
and vertical axis represents frequency or relative frequency.
Figure 2-8. : Histogram
Histograms have certain identifiable characteristics, as shown in Figure 2-8. One characteristic of
the distribution concerns symmetry or lack of symmetry of the data. Is the data equally
distributed on each side of the center of measurement (e.g. Temperature), or it is skewed to right
or left? Another characteristic concerns the kurtosis of the data. A final characteristic concerns
number of modes, or peaks, in the data. There can be one mode, two modes (bi-modal) or
multiple modes.
Histograms can also provide sufficient information about a quality problem to provide a basis for
decision making without statistical analysis. They can also be compared in regard to location,
spread, and shape. A histogram is like a snapshot of the process showing variation in the
characteristic. Histograms can determine process capability, compare with specifications, suggest
shape of the population, and indicate any discrepancies in the data. A typical histogram using
MINITAB software is shown below.
9085807570656055
14
12
10
8
6
4
2
0
Marks in Statistics Course
Freq
uenc
yHistogram
Figure 2-9. : Histogram of Marks in Statistics Course
Run Charts
A run chart, which is shown in Figure 2-10 is a very simple quality tool for analyzing process
with respect to (w.r.t) time in development stage or, for that matter, when other charting
techniques are not quite relevant. The important point is to draw a picture of the process w.r.t.
time and let it "talk" to you. Plotting time oriented data points is a very effective way of
highlighting any pattern observed w.r.t. time. This type of plotting should be done before doing
histogram or any other statistical data analysis.
Figure 2-10: Run Chart
The horizontal axis in Figure 2-10 is labeled as time (Day of the Week), and vertical axis of the
graph represents measurement on variables of interest.
Scatter Diagram
The simplest way to determine if a relationship exists between TWO variables is to plot a scatter
diagram. Figure 2-11 shows a relationship between automotive speed and gas mileage. The
figure indicates that as speed increases, gas mileage decreases or a negative relationship exist
between the variables of interest. Automotive speed is plotted on x-axis and so-called
independent variable. The independent variable is usually controllable. Here, gas mileage is on
the y-axis and is the dependent or so-called response variable.
.
Figure 2-11 Scatter Diagram
There are a few simple steps for constructing a scatter diagram. Data is collected as ordered pairs
(x, y). The automotive speed is controlled and the gas mileage is measured. Horizontal and
vertical scales are constructed with higher values on right for x-axis and on the top for y-axis.
After the scales are labeled, data is plotted. Once the scatter diagram is complete, relationship or
Pearson correlation (http://en.wikipedia.org/wiki/Correlation_coefficient) between two variables
can be found out. In MINITAB, all relevant information can be derived using scatter plot
(GRAPH-Scatter Plot)/correlation/regression option and shown with an example in Figure 2-12,
and Figure 2-13.
950900850800750700650
1400
1300
1200
1100
1000
Chicken Basket (X)
Cold
Dri
nks
Sale
s (Y
)Scatter Plot
Figure 2-12 Scatter Plot in MINITAB with positive correlation of 0.98
3837363534333231
280
260
240
220
200
180
Direction-East
Hea
tFlu
x
Scatter Plot with Week or No Relationship
Figure 2-13 Scatter Plot in MINITAB with week correlation of 0.1
Week correlation will not imply ‘no’ relationship. There may be nonlinear
relationship, which is not reflected by Pearson Correlation Coefficient.
Few other graphical plots extensively used in Quality Data analysis are Control Chart,
and Box Plot. These are discussed below.
Control Chart
Quality control is one approach that any organization adopts to detect defects and to take
corrective actions. Quality control is employed to ensure the desired level of quality in the final
goods and services. Quality control is about analysis of data for rectification of errors with
respect to time. Walter Shewhart developed the control charts in 1924. It focuses on monitoring
the performance of characteristic of interest over a period of time by looking at the variability in
the data. There are two broad categories of control charts: control charts for attributes and control
charts for variables. A variable control chart consists of a centre line (CL) that represents the
mean value of the characteristic of interest. In addition, two other horizontal lines, namely the
Upper Control Limit (UCL) and the Lower Control Limit (LCL), are also shown in the control
chart. A typical variable control chart on mean and range of a characteristic, so-called X-bar and
R is shown below.
2321191715131197531
3.100
3.075
3.050
3.025
3.000
Sample
Sa
mp
le M
ea
n
__X=3.0608
UC L=3.1145
LC L=3.0071
2321191715131197531
0.15
0.10
0.05
0.00
Sample
Sa
mp
le R
an
ge
_R=0.0525
UC L=0.1351
LC L=0
X-bar and R Control Chart for Paint Thickness characteristic
Figure 2-14 A Variable Control Chart
Sample mean ( x ) chart monitors the accuracy and central tendency of the process
output characteristic. Whereas, the sample range ( R ) chart monitor the variation of
the characteristic, with respect to time. The calculation details on UCL, CL and LCL
can be found in any Quality Management Book (Mitra, A, 2008; Montgomery, D.C.,
2008). In attribute type of control chart (used to monitor number of defects or defectives), only
one chart is used to monitor deviation with respect to time. More details on control chart are
given in below section on statistical technique.
Box Plot
Box plot provide a display on quartiles, and outliers for a given data set. If we need to
compare variation of two data set (say two different service time), we may need the
help of Box plot at initial stage before going into inferential statistics and hypothesis
testing. A typical comparative Box Plot for two fast food restaurant is shown below.
KYCMacnalds
350
300
250
200
150
100
Dat
a
Box Plot of Two Restaurant Service Time
Figure 2-15 A Comparative Box Plot
Each box in the graph shows first quartile, second quartile and third quartile. The
extension line (Whisker) beyond the box is minimum of 1.5*(Inter quartile Range)
and extreme data point. ‘*’ beyond the whisker is considered as outlier.
It is observed that although the service time median of Macnalds and KYC seems
close, the variability of KYC data is much more than Macnalds. Thus it seems
Macnalds is more consistent on service time than KYC.
In addition to the above chart, stem and leaf plot
(http://www.youtube.com/watch?v=cOl-d3BERkM) and Multi-vari Chart
(http://en.wikipedia.org/wiki/Multi-vari_chart) are also useful in certain situations.
This two are not discussed in detail and can be found in litratures, books, and web.
In process quality improvement, not only the 7 QC tools and plots are important, but
few statistical techniques are extensively used for inferential statistics and decision
making. Few of them are discussed below.
Statistical Techniques
Few important statistical techniques, frequently used in quality improvement and
decision making, include hypothesis testing, regression analysis, sampling technique,
two sample t-test, Analysis of Variance (ANOVA), and Design of Experiment
(DOE). These techniques are discussed briefly below.
HYPOTHESIS TESTING
Population parameters (say mean, variance) of any characteristic which are of relevance in most
of the statistical studies are rarely known with certainty and thus estimated based on sample
information. Estimation of the parameter can be a point estimate or an interval estimate (with
confidence interval). However, many problems in engineering science and management require
that we decide whether to accept or reject a statement about
some parameter(s) of interest. The statement which is challenged is known as a null hypothesis,
and the way of decision-making procedure
is so-called hypothesis testing. This is one of the most useful techniques
for statistical inference. Many types of decision-making problems in the engineering science can
be formulated as hypothesis-testing problems. If an engineer is interested in comparing mean of
a population to a specified value. These simple comparative experiments are frequently
encountered in practice and provide a good foundation for the more complex experimental
design problems that will be discussed subsequently. In the initial part of our discussion, we will
discuss comparative experiments involving either one or two populations, and our focus is on
testing hypothesis concerning the parameters of the population(s). We now give a formal
definition of a statistical hypothesis.
Definition
A statistical hypothesis is a statement about parameter(s) of one or more
populations.
For example, suppose that we are interested in burning rate of a solid propellant
used to power aircrew escape systems. Now burning rate is a random variable that can
be described by a probability distribution. Suppose that our interest focuses on mean
burning rate (a parameter of this distribution). Specifically, we are interested in deciding
whether or not the mean burning rate is 60 cm/s. We may express this formally as
µµ=
≠0
1
: 60 cm/s: 60 cm/s
HH
The statement µ =0 : 60 cm/sH is called the null hypothesis, and the statement
µ ≠1 : 60 cm/sH is so-called the alternative hypothesis. Since alternative
hypothesis specifies values of µ that could be either greater or less than 60 cm/s, it is
called a two-sided alternative hypothesis. In some situations, we may wish to formulate
a one-sided alternative hypothesis, as
µµ=
<0
1
: 60 cm/s: 60 cm/s
HH
Or µµ=
>0
1
: 60 cm/s: 60 cm/s
HH
It is important to remember that hypotheses are always statements about the population
or distribution under study, AND not statements about the sample. An experimenter generally
believes the alternate hypothesis to be true. Hypothesis-testing procedures rely on using
information in the random
sample from the population of interest. Population (finite or infinite) information is impossible
to collect. If this information is consistent with the null hypothesis then we will conclude that
the null hypothesis is true: however if this information is inconsistent with null hypothesis, we
will conclude that there is little evidence to support null hypothesis.
The structure of hypothesis-testing problems is generally identical in all engineering/science
applications that are considered. Rejection of
null hypothesis always leads to accepting alternative hypothesis. In our treatment of
hypothesis testing, null hypothesis will always be stated so that it specifies an exact
value of the parameter (as in the statement µ =0 : 60 cm/sH ). The
alternative hypothesis will allow the parameter to take on several values (as in the statement
µ ≠1 : 60 cm/sH ). Testing the hypothesis involves taking a random
sample, computing a test statistic from the sample data, and using the test statistic
to make a decision about the null hypothesis.
Testing a Statistical Hypothesis
To illustrate the general concepts, consider the propellant burning rate problem introduced
earlier. The null hypothesis is that the mean burning rate is 60 cm/s. and the alternative is that
it is not equal to 60 cm/s. That is, we wish to test
µµ=
≠0
1
: 60 cm/s: 60 cm/s
HH
Suppose that a sample of n=10 specimens is tested and that the sample mean
burning rate X is observed. The sample mean is an estimate of the true population mean µ .
A value of the sample mean X that falls close to the hypothesized value of µ = 60 /cm s is
evidence that the true mean µ is really 60 cm/s: that is, such evidence supports the null
hypothesis 0H . On the other hand, a sample mean that is considerably different from 60 cm/s is
evidence in support of the alternative hypothesis, 1H . Thus sample mean is the test statistic in
this case.
Varied Sample may have varied mean values. Suppose that if ≤ ≤58.5 61.5x , we will
accept the null hypothesis µ =0 : 60H , and if either < 58.5x or >58.5 61.5x ,
we will accept the alternative hypothesis µ ≠1 : 60H . The values of X that are less than 58.5
and greater than 61.5 constitute the rejection region for the test, while all values that are in the
interval 58.5 to 61.5 forms acceptance
region. Boundaries between critical regions and acceptance region are so-called
‘critical values’. In our example the critical values are 58.5 and 61.5. Thus, we reject Ho in
favor of 1H if the test statistic falls in the critical region and accept Ho otherwise. This interval
or region of acceptance is defined based on the concept of confidence interval and level of
significance for the test. More details on confidence interval and level of significance can be
found in various web link (http://en.wikipedia.org/wiki/Confidence_interval;
http://www.youtube.com/watch?v=iX0bKAeLbDo) and book (Montgomery and Runger,
2010).
Hypothesis decision procedure can lead to either of two wrong conclusions. For example, true
mean burning rate of the propellant could be equal to 60 cm/s. However, for randomly selected
propellant samples that are tested, we could observe a value of the
test statistic, X , that falls into the critical region. We would then reject the null hypothesis 0H
in favor of the alternative, 1H , when, in fact Ho is really true. This type of wrong conclusion is
called a Type I error. This is a more serious mistake as compared to another error Type II
explained below.
Now suppose that true mean burning rate is different from 60 cm/s, yet sample
mean X falls in the acceptance region. In this case we would accept 0H when it is false.
This type of wrong conclusion is called a Type II error.
Thus, in testing any statistical hypothesis, four different situations determine whether
final decision is correct or error. These situations are presented in Table 2- 1.
Because our decision is based on random variables, probabilities can be associated
with the Type I and Type II errors. The probability of making a Type I error
is denoted by Greek letterα . That is,
( ) ( )α = = 0 0Type I error Reject H |H is trueP P
Sometimes the Type I error probability is called the significance level (α) or size of the test.
Table 2-1 Type I and Type II Error
Actual Decision H0 is true H0 is false Accept H0 No error Type II error Reject H0 Type I error No error
The steps followed in hypothesis testing are
(i) Specify the Null and Alternate hypothesis
(ii) Define level of significance(α) based on criticality of the experiment
(iii) Decide the type of test to be used (left tail, right tail etc.)
(iv) Depending on the test to be used, sample distribution, mean variance information, define
the appropriate test statistic (z-test, t-test etc.)
(v) Considering the level of significance (α), define the critical values by looking into standard
statistical tables.
(vi) Decide on acceptance or rejection of null/ alternate hypothesis by comparing test statistic
values calculated from samples, and as defined in step (iv), with standard value specified in
statistical table.
(vii) Derive meaningful conclusion.
Regression Analysis
In many situations, two or more variables are inherently related, and it is necessary to
explore nature of this relationship (linear or nonlinear). Regression analysis is a statistical
technique for
investigating the relationship between two or more variables. For example,
in a chemical process, suppose that yield of a product is related to process operating temperature.
Regression analysis can be used to build a model (response surface) to predict yield at a given
temperature level. This response surface can also be used for further process optimization, such
as finding the level of temperature that maximizes yield, or for process control purpose.
Let us look into Table 2-2 with paired data collected on % Hydrocarbon levels (say x variable)
and corresponding Purity % of Oxygen (say, y variable) produced in a chemical distillation
process. The analyst is interested to estimate and predict the value of y for a given level of x,
within the range of experimentation.
Table 2-2 Data Collected on % Hydrocarbon levels (x) and Purity % of Oxygen (y)
Observation number
Hydrocarbon level x (%)
Purity y (%)
Observation number
Hydrocarbon level x (%)
Purity y (%)
1 0.99 90.1 11 1.19 93.54 2 1.0 89.05 12 1.15 92.52 3 1.15 91.5 13 0.97 90.56 4 1.29 93.74 14 1.01 89.54 5 1.44 96.73 15 1.11 89.85 6 1.36 94.45 16 1.22 90.39 7 0.87 87.59 17 1.26 93.25 8 1.23 91.77 18 1.32 93.41 9 1.55 99.42 19 1.43 94.98
10 1.4 93.65 20 0.95 87.33
Scatter diagram is a firsthand visual tool to understand the type of relationship, and then
regression analysis is recommended for developing any prediction model. Inspection of this
scatter diagram (given in Figure 2-16) indicates that although no simple curve will pass exactly
through all the points, there is a strong trend indication that the points lie scattered randomly
along a straight line.
1.61.51.41.31.21.11.00.90.8
100
98
96
94
92
90
88
86
% Hydrocarbon Level(x)
% P
urit
y (y
)
Scatter Plot and Regression
Figure 2-16 Scatter diagram and Trend of Relationship
Therefore, it is probably reasonable to assume that mean of the random variable Y is related to x
by following straight-line relationship. This can be expressed as
( ) | 0 1| Y xE Y x xµ β β= = +
where, regression coefficient β0 is so-called intercept and β1 is the slope of the line. Slope and
intercept is calculated based on a ordinary least square method. While the
mean of Y is a linear function of x, the actual observed value y does not fall exactly on
a straight line. The appropriate way to generalize this to a probabilistic linear model is
to assume that the expected value of Y is a linear function of x. But for a particular value
of x, actual value of Y is determined by mean value from the linear regression model
plus a random error term,
β β ε= + +0 1Y x
where ε is the random error term. We call this model as simple regression
model, because it has only one independent variable(x) or regressor. Sometimes a model like this
will arise from a theoretical relationship. Many times, we do not have theoretical knowledge of
the relationship between x and y and the choice of the model is based on inspection of a scatter
diagram, such as we did with the oxygen purity data. We then think of the linear regression
model as an empirical model with uncertainty (error).
The regression option in MINITAB can be used to get all the results of the model. The results as
derived from MINITAB-REGRESSION option using oxygen purity data set is provided below.
The regression equation is % Purity (y) = 74.5 + 14.8 % Hydrocarbon Level(x) Predictor Coef SE Coef T P Constant 74.494 1.666 44.73 0.000 % Hydrocarbon Level(x) 14.796 1.378 10.74 0.000 S = 1.13889 R-Sq = 86.5% R-Sq(adj) = 85.7% Analysis of Variance Source DF SS MS F P Regression 1 149.55 149.55 115.30 0.000
Residual Error 18 23.35 1.30 Total 19 172.90
We look into R-Sq value and if it is more than 70%, we assume the relationship is linear and y
depends on x. More details on interpretation of other values are given in MINITAB help menu or
readers can refer to any standard statistical or quality management book.
In this context, P-value and its interpretation is important from the context of hypothesis testing
and regression analysis. General interpretation is that if the P-value is less than 0.05 (at 5% level
of significance test) the NULL HYPOTHESIS is to be rejected. Reader may refer web
(http://www.youtube.com/watch?v=lm_CagZXcv8;
http://www.youtube.com/watch?v=TWmdzwAp88k ) for more details on P-value and its
interpretation.
Common Abuses of Regression
Regression is widely used and frequently misused. Care should be taken in selecting variables
with which to construct regression equations and in determining form of a model. It is possible
to develop statistical relationships among variables that are completely unrelated in a
practical sense, For example, we might attempt to relate shear strength of spot welds
with number of boxes of computer paper used by information systems group. A
straight line may even appear to provide a good fit to the data, but the relationship is an
unreasonable one. A strong observed association between variables does
not necessarily imply that a causal relationship exists between those variables. Designed
experimentation is the only way to prove causal relationships.
Regression relationships are also valid only for values of the regressor variable within
range of original experimental/actual data. But it may be unlikely to remain so as we extrapolate.
That is, if we use values of x beyond the range of observations, we become less certain about the
validity of the assumed model and its prediction.
Process quality monitoring and control
Process quality is monitored by using acceptance sampling technique and control is achieved
through Statistical process control (SPC) chart. Monitoring is essential so as to utilize full
potential of the process. Statistical quality control is dating back to the 1920s. Dr. Walter A.
Shewhart of the Bell Telephone Laboratories was one of the early pioneers of the field. In 1924
he wrote a memorandum showing a modem control chart, one of the basic
tools of statistical process control. Dr. W. Edwards Deming and Dr. Joseph
M. Juran have been instrumental in spreading statistical quality-control methods since
World War II.
In any production process, regardless of how well-designed or carefully maintained it is,
a certain amount of inherent or natural variability will always exist. This natural variability or "
noise" is the cumulative effect of many small, essentially unavoidable causes. When the noise in
a process is relatively small, we usually consider it an acceptable level of process performance.
In the framework of statistical quality control, this natural variability is often called “chance
cause variability". A process that is operating with only chance causes of variation is said to be
in statistical control. In other words the chance causes are an inherent part of the process.
Other kinds of variability may occasionally be present in the output of a process, This variability
in key quality characteristics usually arises from sources, such as improperly
adjusted machines, operator errors, or defective raw materials. Such variability is generally large
when compared to the background noise and it usually represents an unacceptable level of
process performance. We refer these sources of variability that are not part of the chance cause
pattern as assignable causes. A process that is operating in the presence of assignable causes is
said to be out-of-control. There are varied statistical control chart to identify out of control
signal. Typically any control chart will have an upper and lower control limit and a central line
as given in Figure 2-17.
Figure 2-17 Control Chart Limits
There is a close connection between control charts and hypothesis testing. Essentially the control
chart is a test of the hypothesis that the process is in a state of statistical control.
There can be attribute (say monitoring defects or defective) control chart and variable control
chart. Variable control chart example is given earlier. A ‘c’ attribute chart monitors defects. A ‘c’
chart is shown below, where number of defect data in engine assembly are collected over period
of time. Thus at a particular time a sample engine assembly is selected and number of defects in
the assembly is recorded and monitored.
191715131197531
9
8
7
6
5
4
3
2
1
0
Sample
Sam
ple
Coun
t
_C=3.2
UCL=8.5
LCL=0
c Type Control Chart
Figure 2-18 A c- type Attribute Control Chart and limit lines
P-chart is used to monitor defectives.
Details on various types of control chart and their specific application in varied situation can be
seen in many well known text books (Mitra, A, 2008; Montgomery, D.C., 2008) or web
(http://www.youtube.com/watch?v=gTxaQkuv6sU) .
The principles of control chart are based on acceptance sampling plans, provided by Harold F.
Dodge and Harry G. Romig, who are employees of Bell
System. Acceptance sampling plan is discussed in the following section.
Acceptance sampling plan
Acceptance sampling is concerned with inspection and decision making regarding product
quality. In 1930’s and 1940’s, acceptance sampling was one of the major components of quality
control, and was used primarily for incoming or receiving inspection.
A typical application of acceptance sampling is as follows: A company receives a
shipment of product from its vendor. This product is often a component or raw material used in
company's manufacturing process. A sample is taken from the lot, and some quality characteristic
of the units in the sample is inspected referring to specification. On the basis of information in this
sample, a decision is made regarding acceptance or rejection of the whole lot. Sometimes we refer
to this decision as lot sentencing. Accepted lots are put into production; rejected lots are returned
to the vendor or may be subjected to some other lot-disposition action.
Although it is customary to think of acceptance sampling as a receiving inspection activity, there
are other uses of sampling methods. For example, frequently a manufacturer will sample and
inspect its own product at various stages of production. Lots that are accepted are sent forward for
further processing, and rejected lots may be reworked or scrapped.
Three aspects of sampling include:
1) Purpose of acceptance sampling is to take decision on acceptance of lots, not to estimate
the lot quality. Acceptance-sampling plans do not provide any direct form of quality
control.
2) Acceptance sampling simply accepts and rejects· lots. This is a post mortem kind of
activity. Statistical process controls are used to control and systematically improve quality
by reducing variability, but acceptance sampling is not.
3) Most effective use of acceptance sampling is not to "inspect quality into the product," but
rather as an audit tool to ensure that output of a process conforms to requirements.
Advantages and Limitations of Sampling Plan as compared to 100 % inspection
In comparison with 100% inspection, acceptance sampling has following advantages.
(i) It is usually less expensive because there is less inspection.
(ii) There is less handling of product, hence reduced damage.
(iii) It is highly effective and applicable to destructive testing.
(iv) Fewer personnel are involved in inspection activities.
(v) It often greatly reduces amount of inspection error.
(vi) Rejection of entire lots as opposed to the simple return of defectives
provides a stronger motivation to suppliers for quality improvement.
Acceptance sampling also has several limitations which include:
(i) There is risk of accepting "bad" lots and rejecting "good" lots, or Type I and Type II error.
(ii) Less information is usually generated about the product.
Types of Sampling Plans
There are a number of ways to classify acceptance-sampling plans. One major
classification is by attributes and variables. Variables are quality characteristics that are measured
on a numerical scale whereas attributes are quality characteristics are expressed on a "go, no-go"
basis.
A single-sampling plan is a lot-sentencing procedure in which one sample
units is selected at random from the lot, and disposition of the lot is determined
based on information contained in that sample. For example, a single-sampling
for attributes would consist of a sample size n and an acceptance number c. The procedure would
operate as follows: Select n items at random from the lot. If there are
fewer than c defectives in the sample, accept the lot, and if there are more than c defective in the
sample, reject the lot.
Double-sampling plans are somewhat more complicated. Following an initial sample, a decision
based on the information in that sample is made either to accept
the lot, reject the lot or to take a second sample. If the second sample is taken,
information from both first and second sample is combined in order to reach a
decision whether to accept or reject the lot.
A multiple-sampling plan is an extension of the double-sampling concept, in that
more than two samples may be required in order to reach a decision regarding the disposition of
the lot. Sample sizes in multiple sampling are usually smaller than they are in either single or
double sampling. The ultimate extension of multiple sampling is sequential sampling, in which
units are selected from the lot one at a time, and following
inspection of each unit, a decision is made either to accept the lot, reject the lot, or select
another unit.
Random Sampling
Units selected for inspection from the lot should be chosen at random, and they
should be representative of all the items in the lot. Random-sampling concept is extremely
important in acceptance sampling and statistical quality control. Unless random samples are used,
bias may be introduced. Say, suppliers may ensure that units packaged on the top of the lot are of
extremely good quality, knowing that inspector will select sample
from the top layer. This also helps in identifying any hidden factor during experimentation.
The technique often suggested for drawing a random sample is to first assign a
number to each item in the lot. Then n random numbers are drawn (from random number table or
using excel/statistical software), where the range of
these numbers is from 1 to the maximum number of units in the lot. This sequence of
random numbers determines which units in the lot will constitute a sample. If products have serial
or other code numbers, these numbers can be used to avoid process
of actually assigning numbers to each unit. Details on different sampling plan can be seen in
Mitra, A (2008).
Process improvement Tools
Acceptance sampling and statistical quality control techniques may not significantly reduce
variability in the output. Process improvement by variation reduction is an important feature in
quality management. There are varieties of statistical tools available for improving processes.
Some of them are discussed below.
ANOVA
Many experiments involve more than two levels of a factor. Experimenter is interested to
understand the influence of the factor on variability of output characteristic. In this case,
Analysis of variance (ANOVA) is the appropriate statistical technique. This technique is
explained with the help of following example.
Say, a product development engineer is interested in investigating tensile strength of a new
synthetic fiber. The engineer knows from previous experience that the strength is affected by
weight percent of cotton used in the blend of materials for the fiber. Furthermore, he suspects that
increasing cotton content will increase the strength, at least initially. He also knows that cotton
content should range of 1 to 25 percent if final product is to have other quality characteristics
that are desired. Engineer decides to test specimens at five levels of cotton weight percent: 5, 10,
15, and 20 percent. She also decides to test five specimens at each level of cotton content. This
is an example of a single-factor (Cotton Weight %) experiment with a (level) = 5 of the factor
and n = 5 replicates. The 25 runs should be made in random sequence.
Table 2-3 Experimental Data
Experimental run number Cotton weight percentage 1 2 3 4 5
5 7 8 15 11 9 10 12 17 13 14 19 15 14 18 19 17 16 20 25 22 23 18 20
The randomized test sequence is necessary to prevent the effects of unknown nuisance variables
(or hidden factor influence), perhaps varying out of control during experiment, from
contaminating the results. Balanced experiments with equal number of replicates are also preferred
to minimize the experimental error.
This is so-called a single-factor analysis of variance model and known as fixed effects model
as the factor level can be changed as required by the experimenter . Recall that yij, represents the
jth sample (j=1,..n, replicate) output observations under ith treatment combination. Let iy •
represent the average of the observations under the ith treatment. Similarly, let y•• represent the
grand total of all the observations and represent the grand average of all the observations.
Expressed symbolically,
1
1 1
i=1,2,...,a
n
i ij i ij
a n
ij ii i
y y y y n
y y y y N
=
⋅⋅ ⋅⋅= =
= =
= =
∑
∑∑
where, N or 𝑎𝑎 ∗ 𝑛𝑛 is the total number of observations. The "dot" subscript notation used in
above equations implies summation over the subscript that it replaces.
The appropriate hypotheses are
0 1 2
1
: 0: for at least one pair (i, j)µ µ µµ µ
= = = =
≠
a
i j
HH
Decomposition of the total sum of squares
The name analysis of variance is derived from a partitioning of total variability into its
component parts. The total corrected sum of squares
( )2
1 1
a n
T iji j
SS y y⋅⋅= =
= −∑∑
is used as a measure of overall variability in the data. Intuitively, this is reasonable because, if we
were to divide SS, by the appropriate number of degrees of freedom (in this case, N -1), we would
have the sample variance of the y 's. The sample variance is of course a standard measure of
variability.
Note that the total corrected sum of squares SST may be written as
( ) ( )( )2
2
1 1 1 1
a n a n
T ij i ij ii j i j
SS y y y y y y⋅⋅ ⋅ ⋅⋅ ⋅= = = =
= − = − − ∑∑ ∑∑
or,
( ) ( ) ( ) ( )( )2 22
1 1 1 1 1 1 12
a n a a n a n
T ij i ij i i ij ii j i i j i j
SS y y n y y y y y y y y⋅⋅ ⋅ ⋅⋅ ⋅ ⋅ ⋅⋅ ⋅= = = = = = =
= − = − + − + − −∑∑ ∑ ∑∑ ∑∑
and as it is proved that the third product term vanishes, we can rewrite the overall expression of
SSTotal (SST) as
= +T Treatments ESS SS SS
Where, SSTreatments, is so-called the sum of squares due to treatments (i.e., between
treatments), and SSError, is called the sum of squares due to error (i.e., within treatments). There
are an N total observations: thus SST has N - 1 degrees of freedom. If there area levels of the
factor (and a treatment means), so SSTreatments, has a - 1 degrees of freedom. Finally, within
any treatment there are n replicates providing n - 1 degrees of freedom with which to
estimate the experimental error. Because there are a treatments, we have a(n - 1) = an - a = N -
a degrees of freedom for error.
Statistical Analysis
The Analysis of variance table (Table 2-4) for the single-factor fixed effects model is given
below
Table 2-4 ANOVA Table with formula
Source of
variance
Sum of squares Degrees of
freedom
Mean square F0
Between
treatments 2
1( )
Treatmentsa
ii
SS
n y y• ••=
= −∑
a-1 MSTreatments 0
Treatments
E
MSF
MS=
Error (within
treatments)
= −Error T TreatmentsSS SS SS
a(n-1) MSE
Total 2
1 1( )
a n
T iji j
SS y y••= =
= −∑∑
an-1
Because the degrees of freedom for TreatmentsSS and SSError add to N-1, the total number of
degrees of freedom, Cochran’s theorem implies that 2/TreatmentsSS σ and 2/ESS σ are
independently distributed chi-square random variables. Therefore, if the null hypothesis of no
difference in treatment means is true, the ratio
( )( )
/ 1/
Treatments Treatmentso
E E
SS a MSF
SS N a MS−
= =−
,
is distributed as F with a - 1 and N - a degrees of freedom. Equation above is the test statistic
for the hypothesis of no differences in treatment means.
From the expected mean squares we see that, in general, EMS is an unbiased estimator of 2σ .
Also, under the null hypothesis, TreatmentsMS is an unbiased estimator of 2σ . However, if the null
hypothesis is false, the expected value of TreatmentsMS is greater than 2σ . Therefore, under the
alternative hypothesis, the expected value of the numerator of the test statistic (Equation given
above for Fo) is greater than the expected value of the denominator, and we should reject Ho
on values of the test statistic that are too large. This implies an upper-tail and one-tail critical
region. Therefore, we should reject Ho and conclude that there are differences in the treatment
means if
, 1,o a N aF Fα − −> ,
where, Fo is computed from above equation. Alternatively, we can also use the p-value approach
for decision making as provided by statistical softwares, say MINITAB, SAS.
Using MINITAB, we can obtain the following graphs and results for the above mentioned
experiment on tensile strength:
2015105
25
20
15
10
5
Weight percent of cotton
Tens
ile S
tren
gth
Boxplot of Tensile Strength
Figure 2-19 Box Plot of Data
From Box-plot it is observed that as cotton weight % increases tensile strength also improves.
However, whether any two means are significantly different cannot be commented based on
Box plot.
5.02.50.0-2.5-5.0-7.5
99
95
90
80
70
60504030
20
10
5
1
Residual
Perc
ent
Normal Probability Plot(response is Tensile Strength)
Figure 2-20 Residual Plot The residual plot confirms Normality assumption of error. Conclusion in case error is non-
normal may be erroneous. Normality assumption can be tested by using Anderson-Darling test
statistic value provided in MINITAB (Stat->Basic Statistics-> Normality Test). One-way ANOVA Analysis: Tensile Strength versus Weight % of cotton Source DF SS MS F P Weight percent of cotton 3 340.15 113.38 14.77 0.000 Error 16 122.80 7.67 Total 19 462.95 S = 2.770 R-Sq = 73.47% R-Sq(adj) = 68.50% Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev -----+---------+---------+---------+---- 5 5 10.000 3.162 (----*----) 10 5 15.800 3.114 (-----*----) 15 5 16.800 1.924 (-----*----) 20 5 21.600 2.702 (----*----) -----+---------+---------+---------+---- 10.0 15.0 20.0 25.0
Pooled St. Dev = 2.770
Figure 2-21 ANOVA Results in MINITAB
ANOVA results given above confirm that changing the % weight influences tensile
strength and it is linear (from R-square value).
For determining the best setting of % cotton weight, one can do the Fisher LSD
comparison test as given in Figure 2-22.
Fisher 95% Individual Confidence Intervals All Pairwise Comparisons among Levels of Weight percent of cotton
Simultaneous confidence level = 81.11% Weight percent of cotton = 5 subtracted from: Weight percent of cotton Lower Center Upper ----+---------+---------+---------+----- 10 2.086 5.800 9.514 (----*-----) 15 3.086 6.800 10.514 (-----*----) 20 7.886 11.600 15.314 (-----*----) ----+---------+---------+---------+----- -7.0 0.0 7.0 14.0 Weight percent of cotton = 10 subtracted from: Weight percent of cotton Lower Center Upper ----+---------+---------+---------+----- 15 -2.714 1.000 4.714 (----*-----) 20 2.086 5.800 9.514 (----*-----) ----+---------+---------+---------+----- -7.0 0.0 7.0 14.0
Weight percent of cotton = 15 subtracted from:
Weight percent of cotton Lower Center Upper ----+---------+---------+---------+----- 20 1.086 4.800 8.514 (----*----) ----+---------+---------+---------+----- -7.0 0.0 7.0 14.0
Figure 2-22 Fisher Comparison Test
The comparison tests confirm that cotton weight % 20 is significant ly different
from 15 %, and thus setting of 20 is suggested. Details on interpretat ion of
comparison test and ANOVA is given in MINITAB example, MINITAB help, and
any standard text book on Quality (Montgomery, D. C., 2014)
Designed Experiment
Statistical design of experiments refers to the process of planning an experiment so that
appropriate data which can be analyzed by statistical methods can be collected, resulting in
valid and objective conclusions. Statistical approach to experimental design is necessary if we
wish to draw meaningful conclusions from the data. This helps in confirming any causal
relationship. When problem involves data that are subject to experimental errors, statistical
methodology is the only objective approach for analysis. Thus, there are two aspects to any
experimental problem: design of experiment and statistical analysis of data.
Three basic principles of experimental design are replication, randomization and blocking
(local control). By replication we mean a repetition of basic trial on different sample. In the met-
allurgical experiment, twice replication would consist of treating two specimens by oil
quenching. Thus, if five specimens are treated in a quenching medium in different time point, we
say that five replicates have been obtained.
Randomization is the cornerstone underlying use of statistical methods in experimental design.
By randomization we mean that both allocation of experimental material and order in which
individual runs or trials of experiment are to be performed are randomly determined. Statistical
methods require that observations (or errors) be independently distributed random variables.
Randomization usually validates this assumption. By properly randomizing an experiment, we also
assist in "averaging out" the effects of extraneous (hidden) factors that may be present.
Blocking (or local control) nuisance variables is a design technique used to improve precision
with which comparisons among factors of interest are made. For example, an experiment in a
chemical process may require two batches of raw material to make all required runs. However,
there could be differences between batches due to supplier-to-supplier variability, and if we are
not specifically interested in this effect, we would think of different batches of raw material as a
nuisance factor. Generally, a block is a set of relatively homogeneous experimental conditions.
There are many design options for blocking nuisance variables.
Guidelines for designing experiment
To use statistical approach in designing and analyzing an experiment, it is necessary for
everyone involved in the experiment to have a clear idea in advance of exactly what is to be
studied (objective of study), how data is to be collected, and at least an understanding of how
this data is to be analyzed. Below section briefly discusses on outline and elaborate on some of
the key steps in Design of Experiment (DOE). Remember that experiment can fail. However, it
always provides some meaningful information.
Recognition of a problem and its statement-This may seem to be rather obvious point, but
in practice it is often not simple to realize that a problem requiring experimentation exists, nor
is it simple to develop a clear and generally accepted statement of problem. It is necessary to
develop all ideas about objectives of experiment. Usually, it is important to solicit input from all
concerned parties: engineering, quality assurance, manufacturing, marketing, management,
customers (internal or external) and operating personnel.
It is usually helpful to prepare a list of specific problems or questions that are to be addressed by
the experiment. A clear statement of problem often contributes substantially to better
understanding of phenomenon being studied and final solution to the problem. It is also
important to keep overall objective in mind.
Choice of factors, levels, and range- When considering factors that may influence performance
of a process or system, experimenter usually finds that these factors can be classified as either
potential design (x) factors or nuisance (z) factors. Potential design factors are those factors that
experimenter may wish to vary during the experiment. Often we find that there are a lot of potential
design factors, and some further classification of them is necessary. Some useful classification is
design factors, held-constant factors, and allowed to-vary factors. The design factors are the
factors actually selected for study in the experiment. Held-constant factors are variables that may
exert some effect on the response, but for purposes of present experiment these factors are not of
interest, so they will be held at a specific level.
Nuisance (allowed to-vary) factors, on the other hand may have large effects that must be
accounted for, yet we may not be interested in them in the context of the present experiment.
Selection of the response variable- In selecting response variable, experimenter should be
certain that this variable really provides useful information about process under study. Most
often, average or standard deviation (or both) of the measured characteristic will be response
variable. It iscritically important to identify issues related to defining responses of interest
and how they are to be measured before conducting the experiment. Sometimes designed
experiments are employed to study and improve the performance of measurement systems.
Choice of experimental design- If the pre-experimental planning activities mentioned
above are done correctly, this step is relatively easy. Choice of design involves consideration of
sample size (number of replicates) keeping in mind precision required for experiment, selection of
a suitable run order for the experimental trials and determination of whether or not blocking
restrictions are to be involved.
Performing the experiment- When running an experiment, it is vital to monitor the process
carefully to ensure that everything is being done according to plan. Errors in experimental
procedure or instrumental error during measurement at this stage will usually destroy
experimental validity. Up-front planning is crucial to success. It is easy to underestimate the
logistical and planning aspects of running a designed experiment in a complex manufacturing or
research and development environment. Coleman and Montgomery (1993) suggest that prior to
conducting experiment a few trial runs or pilot runs are often helpful.
Statistical analysis of the data-Statistical methods should be used to analyze the data so that
results and conclusions are objective rather than judgmental in nature. If the experiment has
been designed correctly statistical methods required are not elaborate. There are many excellent
software packages (JMP, MINITAB, and DESIGN EXPERT) to assist in data analysis. Often we
find that simple graphical methods play an important role in data analysis and interpretation.
Conclusions and recommendations- Once the data is analyzed, experimenter must draw
practical conclusions from the results, and recommend a course of action. Graphical methods
are often useful in this stage, particularly in presenting results to others. Follow-up runs and
confirmation testing should also be performed to validate the conclusions from the experiment.
There are many design options for statistical experiment. For two factor experiment, the basic
design is a two-way ANOVA. For higher number of factors, factorial design using orthogonal
array is typically used. There are also central composite designs (CCD), extremely useful to
identify higher order terms in the response surface model developed based on factorial design
(http://www.youtube.com/watch?v=Z-uqadwwFsU) . Three level Box–Behnken or BBD design
(http://www.itl.nist.gov/div898/handbook/pri/section3/pri3362.htm) is also very useful in
situation to identify quadratic terms and interaction in factorial design. Fractional factorial design
is recommended if more than 8 factors are to be studied and there is a need to reduce the number
of factors. This is also known as ‘screening experiment’. Sequential DOE or Response surface
design is used to reach to global optimal setting in case of unimodal function. Taguchi’s method
is a nonconventional approach used when there is little or no higher order interaction.
Desirability function and dual response optimization may be used in case there are multiple y’s
to be optimized simultaneously. Book by Montgomery, D.C. (2014) is an excellent reference to
learn DOE techniques.
Module II. Process Quality Improvement
Lecture – 3 How TQM principal is aligned with process quality improvement?
Total quality management (TQM) is a strategy for implementing and managing quality improvement
activities on an organization-wide basis. TQM began in early 1980s, with the philosophies of Edward
Deming and Joseph Juran as the focal point. It evolved into a broader spectrum of concepts and
ideas, involving participative organizations and work culture, customer focus, supplier quality
improvement, integration of quality system with business goals, and many other activities to focus all
elements of organization around quality and process improvement goal. Typically, organizations that
have implemented TQM approach to quality improvement have quality councils or high-level teams
that deal with strategic quality initiatives, workforce-level teams that focus on outline production or
business activities, and cross-functional teams that address specific process quality improvement
issues. TQM strongly emphasizes on variability reduction, the prime theme of process quality
improvement.
However, TQM has only moderate success for process improvement for a variety of reasons. Some
general reasons for lack of conspicuous success of TQM include (i) lack of top down, high-level
management commitment and involvement; (ii) inadequate use of statistical methods and insufficient
recognition of variability reduction as a prime objective; (iii) diffuse as opposed to focused, specific
objectives; and (iv) too much emphasis on widespread training as opposed to focused technical
education and actual implementation.
Some of the approaches which is extended and used from TQM philosophies are:
Quality Standards
International Standards Organization (ISO) has developed a series of quality standards including ISO
9000 series. Focus of these standards is quality system, including components such as management
responsibility for quality, design control, document and data control, purchasing and contract
management, product identification and traceability, inspection and testing, including control of
measurement and inspection equipment, process control, handling of nonconforming product,
corrective and preventive actions ,handling, storage, packaging, and delivery of product, service
activities, control of quality records ,internal audits, training and statistical methods.
Just-in-Time, Lean Manufacturing, Poka-Yoke, and Others
There are many initiatives devoted for improving production/manufacturing process. Some of
these include Just-in-Time approach emphasizing in-process inventory reduction, reduced set-up
time (SMED), and a pull-type production system; Poka-Yoke or mistake-proofing in the
processes; Toyota production system (TPS), reengineering; theory of constraints (TOC); agile
manufacturing; lean manufacturing; and so on.
Customer Focus
Most important asset of any organization is its customers. An organization's success
depends on how many customers it has, how much they buy, and how often they buy.
Satisfied customers buy more and more frequently. They also pay their bills promptly, which
greatly improves cash flow of an organization.
Increasingly, manufacturing and service organizations are using customer satisfaction as a
measure of quality. This fact is reflected in Malcolm Baldrige National Quality Award, where
customer satisfaction accounts for 30 percent of the total points. Similarly, customer satisfaction
standards are woven throughout the ISO 9000: 2008 standard. Customer satisfaction is one of
the major focuses of a effective quality management system.
Who are the Customers?
There are two distinct types of customers-external and internal. An external customer is the one
who uses the end product or service or one who purchases product or service. Whereas, internal
customer are always defined within the organization/interlinked processes (e.g. marketing
department may be an internal customer for production department).
Customer Perception of Quality
One of the basic concepts of TQM philosophy is continual improvement. This concept implies
that there is no acceptable quality level because the customer's needs, values, and expectations
are constantly changing.
Before making a major purchase, some people check consumer magazines that rate product
quality. During 1980 to 1988, quality of a product and its performance ranked first, price was
second, and service was third. During 1989 to 1992, product quality remained the most
important factor, but service is ranked above price in importance.
An American Society for Quality (ASQ) survey on end user perceptions for important factors
that influenced purchases showed following ranking:
1. Performance
2. Features
3. Service
4. Warranty
5. Price
6. Reputation
The factors of performance, features, service, and warranty are part of the product or
service quality; therefore, it is evident that product quality and service are more important than
price. Although this information is based on retail customer, it appears, to some extent, to be true
for the organizational customer also.
Translating the Customer Needs into Requirements
Kano model, which is shown in Figure 2-23, conceptualizes customer requirements. The model
represents three major areas of customer satisfaction. The first area of customer satisfaction,
represented by diagonal line, represents explicit requirements. These include written or verbal
requirements and are easily identified, expected to be met, and typically performance related.
Satisfying customer would be relatively simple if these were the only requirements.
The second area of customer satisfaction represents innovations, as shown by
curved line in the upper left corner of the figure. A customer's written instructions are
often purposefully vague to avoid stifling new ideas during conceptualization and product
definition. Because they are unexpected, these creative ideas often excite and delight
the customer. These ideas quickly become expected w.r.t time.
The third and most significant area of customer satisfaction represents unstated or unspoken
requirements, as shown by the curve in the lower right corner of the Figure.
Figure 2-23 The Kano Model
The customer may indeed be unaware of these requirements, or they may assume that such
requirements will be automatically supplied. Basic specifications often fail to take real-world
manufacturing requirements into account; many merely are based on industry standards or past
practice. These implied requirements are most difficult to define but prove very costly if
ignored. They may be rediscovered during an after-the- fact analysis of lessons learned.
Realistically customer doesn’t buy a specification; customers buy a product or service to fulfill
his need. Peter Drucker once said, “Customers don’t buy products, they buy results”. Just
meeting a customer’s needs is not enough; organization must exceed customer’s needs.
Questioners are designed to identify basic, expected and exiting features according to Kano’s
Model and then translated into engineering requirement by using QFD.
Module II. Process Quality Improvement
Lecture – 4 How leadership influence process quality initiatives?
There is no universal definition of leadership and indeed many books have been devoted to the
topic of leadership. Researchers describe a leader as one who instills purposes, not one who
controls by brute force. A leader strengthens and inspires followers to accomplish shared goals.
Leader shapes, promotes, protects and exemplifies organization's values. Similarly, Daimler
Chrysler's CEO, Bob Eaton, defines a leader as "someone who can take a group of people to a
place they don't think they can go." As above definitions illustrate, leadership is difficult to
define in anything other than lofty words. The Malcolm Baldrige National Quality Award has a
more grounded definition of leadership in its core values. As stated in its core values and
concepts, visionary leadership is,
"An organization's senior leaders should set directions and create a customer focus, clear
and visible values, and high expectations. Directions, values, and expectations should
balance needs of all stakeholders. Leaders should ensure creation of strategies, systems, and
methods for achieving excellence, stimulating innovation, and building
knowledge and capabilities. Values and strategies should help guide all activities and
decisions of organization. Senior leaders should inspire and motivate entire workforce and
should encourage all employees to contribute, develop and learn, be innovative, and creative.
Senior leaders should serve as role models through their ethical behavior and their personal
involvement in planning, communication, coaching, development of future leaders, review of
organizational performance, and employee recognition. As role models, they can reinforce values
and expectations while building leadership, commitment, and initiative throughout your
organization."
Although leadership is difficult to define, successful quality leaders tend to have certain
characteristics as-
They give priority to external and internal customers and their needs.
Empower, rather than control, subordinates
Emphasize improvement rather than maintenance
Emphasize prevention.
Encourage collaboration rather than competition.
Train and coach, rather than direct and supervise.
Learn from problems.
Continually try to improve communication.
Demonstrate their commitment to quality.
Choose suppliers on the basis of quality, not price.
Establish organizational systems to support the quality effort.
Encourage and recognize team effort.
In order to improve process and system, leadership requires an intuitive understanding of human
nature-basic needs, wants, and abilities. To be effective, a leader understands that people
paradoxically need security and independence at the same time.
Leaders need to give their employees independence and yet provide a secure working
environment-one that encourages and rewards successes. A working environment must
be provided that fosters employee creativity and risk-taking by not penalizing mistakes. This is a
key part for process quality improvement.
A leader will focus on a few key values and objectives in the process. Focusing on a few values
or objectives gives the employees the ability to discern on a daily basis what is important and
what is not in the process. Employees, upon understanding the objectives, must be given
personal control over the process in order to make the task their own and, thereby, something to
which they can commit. A leader, by giving the employee a measure of control over an
important process, will tap into the employee's inner drive. Employees, led by the manager can
become excited participants in the organization.
Having a worthwhile cause such as total quality management is not always enough
to get employees to participate in process improvement. People follow a leader, not a
cause. If the leader is trusted and liked, then employees will participate in total quality
management causes.
Therefore, it is particularly important that a leader's character and competence, which is
developed by good habits and ethics, be above reproach. Effective leadership for improvement
begins on inside and moves out.
Ethics
Ethics is not a precept that is mutually exclusive from quality. Indeed, quality and ethics
have a common care premise, which is to do right things right.
Ethics is a body of principles or standards of human conduct that governs behavior
of individuals and organizations. It is about knowing what the right thing is. Ethics can mean
something different to different people, especially given
an organization's international workforce and varying cultural norms. Because individuals have
different concepts of what is right, leaders need to develop
standards or code of ethics. Quality is dependent on ethical behavior. Doing what is right in the
first place is a proven
way to reduce costs, improve process quality, and create higher customer satisfaction. Many
companies also hire ethics consultant to help them achieve their goal towards improvement.
Core Values, Concepts and Framework
Unity of purpose is the key to a leadership system. Core values and concepts provide that
unity of purpose. Core values and concepts enable a framework for leaders through-
out the organization to make right decisions. They foster TQM behavior and define
culture. Each organization needs to develop its own values. Given below are few core
values, concepts, and framework from Malcolm Baldrige National Quality Award.
They may be used as a starting point for any organization for quality improvement initiatives.
Visionary leadership
An organization's senior leaders need to set directions and create a customer orientation,
clear and visible quality values, and high expectations. Values, directions, and expectations need
to address all stakeholders. The leaders need to ensure the creation of strategies, systems, and
methods for achieving process excellence. Strategies and values should help guide all activities
and decisions of the organization. The senior leaders must commit to the development of the
entire workforce and should encourage participation, learning, innovation, and creativity by all
employees. Through their personal roles in planning, communications, review or organization
performance, and employee recognition, the senior leaders serve as role models, reinforcing the
values and expectations, and building leadership and initiative throughout the organization.
Customer-Driven Excellence
Quality is judged by its customers. All product and service characteristics that contribute
value to the customer and lead to customer satisfaction is the focus of an organization's process
management system. Customer-driven excellence
has both current and future components: understanding today's customer desires and
marketplace offerings as well as future innovations. Value and satisfaction may be influenced by
many factors throughout the customer's overall purchase, ownership, and
service experiences. These factors include the organization's relationship with customers that
helps build trust, confidence, and loyalty. This concept of quality includes
not only the product and service characteristics that meet basic customer requirements, but it also
includes those features and characteristics that differentiate them
from competing offerings. Customer-driven quality is thus a strategic concept. It is directed
toward customer retention, market-share gain, and growth. It demands constant sensitivity to
changing the process and emerging customer and market requirements and the factors that drive
customer satisfaction and retention. It also demands awareness of developments in technology
and of competitors' offerings, and rapid and flexible responses to customer and market
requirements.
Agility
Success in global markets demands agility. Organizations face ever-shorter cycles for
introduction of new and improved products and services, as well as for faster
and more flexible response to customers. Major improvements in response time often
require simplification of work units and processes and ability for rapid changeover
from one process to another. Cross-trained and empowered employees are vital assets in
such a demanding environment.
Managing for Innovation
Innovation means making meaningful change to improve an organization's products,
services and processes to create value for organization’s stakeholders. Innovation can lead an
organization to new dimensions of performance. Innovation is
no longer strictly the purview of research and development departments; innovation is important
for all aspects of business process. Organizations should be
led and managed so that innovation becomes part of organization culture.
Management by Fact
Organizations depend on measurement and analysis of process performance. Such measurements
should derive from business needs and strategy, and should provide critical data and information
about key processes, outputs, and results. Performance measurement should include customer,
product, and service performance; comparisons of operational, market, and competitive
performance; and supplier, employee, and cost and financial performance.
Systems Perspective
The Baldrige Criteria provide a systems perspective for managing an organization to
achieve performance excellence. The Core Values form building blocks and integrating
mechanism for the system. However, successful management of overall performance requires
organization-specific synthesis and alignment. Synthesis means
looking at an organization as a whole and building upon key business requirements, including
strategic objectives and action plans. Alignment means using key linkages
among requirements given in Baldrige Categories, including key measures/
indicators. Alignment includes using measures/indicators to link key strategies with
key processes and align resources to improve overall performance to satisfy customers.
Thus, a systems perspective means managing whole organization, as well as its
components to achieve success.
More details on Baldrige Categories, leadership and quality leadership can be seen in the book
by Besterfield et al. (2004), Evans (2005).
Module II. Process Quality Improvement
Lecture -5 How Lean and JIT are aligned with quality philosophy?
Just-in-time is a philosophy of continual improvement. Lean production process means
supplying the customer with exactly what the customer wants, when the customer wants it, without
waste, through continual improvement. Lean production is driven by "pull" system of the customer's
order. JIT is one of the key ingredient of lean production. When implemented comprehensive
manufacturing strategy, JIT and lean production sustain competitive advantage and result in greater
overall returns.
With JIT, components are "pulled" through the system to arrive where they are needed and
when they are needed. When units do not arrive just as needed, a "problem" is identified. This
makes JIT an excellent tool to help operations managers add value by driving out waste and
unwanted variability. Because there is no excess inventory or excess time in a JIT system, costs
associated with unneeded inventory are eliminated and throughput is improved. Consequently,
the benefits of JIT are particularly helpful in supporting strategies of rapid response at lower cost.
As elimination of waste and variability are fundamental to both JIT and lean production, a brief
explanation on both is provided below.
Waste: It is anything that does not add value to customers. In other words, customers are not
willing to pay for it. Products being stored, inspected or delayed, products waiting in queues, and
defective products which do not add value are waste. Moreover, any activity that does not add
value to a product from the customer's perspective is called waste. JIT provides faster delivery,
reduces work-in-process, and speedy throughput. Additionally, because JIT reduces work-in-process,
it provides little room for any errors, putting added emphasis on quality production. These waste
reduction efforts improves productivity and processes.
Variability Reduction: To achieve just-in-time material movement, managers reduce variability
caused by both internal and external factors in the process. Variability is any deviation in the
standard process to deliver perfect product on time, every time. Reducing inventory, less waste in
the system will ultimately reduce uncertainty. Most variability is caused by tolerating waste or by
poor inventory management or ineffective process. Variability occurs because:
(i) Employees. machines, and suppliers produce units that do not conform to standards,
(ii) Engineering drawings or specifications are inaccurate. It is rare event although.
(iii) Customer’s exact demands are unknown and improper design.
Variability can often go unseen when inventory exists. JIT philosophy is aligned with continual
improvement by reducing such variability. The removal of variability allows us to move
materials just-in-time for use. J1T implementation can reduce throughput time in a supply chain.
Quality improves as uncertainty decreases and variability reduces.
Module II. Process Quality Improvement
Lecture -6 How benchmarking helps to improve process quality?
Benchmarking is a systematic method by which organizations can measure themselves against
the best industry practices. It promotes superior performance by providing an organized
framework through which organizations learn how the "best in class" can do things, understand
how these best practices differ from their own and implement change to close the gap.
Benchmarking Defined
Benchmarking is the systematic search for best practices, innovative ideas, and highly effective
operating practices. Benchmarking considers experience of others and uses it. Indeed, it is a
common-sense proposition to learn from others what they do right and then imitate it to avoid
reinventing the wheel. Benchmarking is not new and indeed has been around for a long time.
Infact, in the 1800s, Francis Lowell, a New England colonist, studied British textile mills and
imported many ideas along with improvements he made for the burgeoning American textile
mills. Benchmarking is used extensively by both manufacturing and service organizations,
including Xerox, AT&T, Motorola, Ford, and
Toyota. Benchmarking is a common element of quality standards, such as the
Chrysler, Ford, and General Motors Quality System Requirements.
Figure 2-24 Benchmarking Framework
As shown in Figure 2-24, benchmarking measures performance against that of best-in- class
organizations (may be entirely different organization with different product or service),
determines how the best in class achieve those performance levels, and uses the information as
the basis for adaptive creativity and breakthrough performance.
Implicit in the definition of benchmarking are two key elements. First, measuring performance
requires some sort of units of measurement. These are called metrics and are usually expressed
numerically. The numbers achieved by the best-in-class are assumed benchmark or target. An
organization seeking improvement then plots its own performance against the target. Secondly,
benchmarking requires that managers understand why their performance differs. Benchmarkers
must develop a thorough and in-depth knowledge of both their own processes and the processes
of the best-in-class organization. An understanding of the differences allows managers to
organize their improvement efforts to meet the desired goal. Benchmarking is all about meeting
goals and objectives by improving processes.
Reasons to Benchmark
Benchmarking is a tool to achieve business and competitive objectives. It is powerful and
extremely effective when used for the right reasons and aligned with organization strategy. It is
not a panacea that can replace all other quality efforts or management processes. Organizations
must still decide which markets to serve and determine the strengths that will enable them to gain
competitive advantage. Benchmarking is one tool to help organizations develop those strengths
and reduce their weaknesses.
By definition, benchmarking requires an external orientation, which is critical in a competitive
world where the competitor can easily be on the other side of the globe. An external outlook
greatly reduces the chance of being caught unaware by the competition. Benchmarking can
notify the organization if it has fallen behind the competition or failed to take advantage of
important operating improvements developed elsewhere. In short, benchmarking can inspire
managers (and organizations) to compete. The primary weakness of benchmarking, however, is
the fact that best-in-class performance is a moving target.
Pitfalls and Criticisms of Benchmarking
The basic idea of benchmarking can be summed up quite simply. Find someone who executes a
process better than you do and imitate what he or she does. The most persistent criticism of
benchmarking comes from the idea of copying others. How can an organization be truly superior
if it does not innovate to get ahead of competitors? It is a question, but one can also ask the
reverse: How can an organization even survive if it loses track of its external environment?
Benchmarking is not a strategy, nor is it intended to be a business philosophy. It is an
improvement tool and must be used properly. Benchmarking isn't very helpful if it is used for
processes that don't offer much opportunity for improvement. It breaks down if process owners
and managers feel threatened or do not accept and act on findings. Over time, things change and
what was state-of-the-art yesterday may not be today. Some processes may have to be
benchmarked repeatedly.
Benchmarking is also not a substitute for innovation; however, it is a source of ideas from
outside organization. Benchmarking forces an organization to set goals and objectives based on
external reality. Consumers care about quality, cost, and delivery, and not productivity of the
organization.
Module II. Process Quality Improvement
Lecture – 7 In what way failure mode and effect analysis (FMEA) helps process
improvement initiative?
Like Design FMEA, as discussed elaborately in Module 3, process FMEA can also be performed
to remove any possibility of process failures. All failure modes, causes of failure(s), severity,
occurrence and detection rating are to be determined and reduction of RPN number by taking
corrective action is the standard procedure of preparing and following Process FMEA. More
discussion is given in QS 9000 (http://en.wikipedia.org/wiki/QS9000) document on process
FMEA (http://asq.org/learn-about-quality/process-analysis-tools/overview/fmea.html;
http://www.qualitytrainingportal.com/resources/fmea/fmea_process.htm). Reader may refer
Module 3 to understand the basic steps of FMEA and rating system.
Module II. Process Quality Improvement
Lecture – 8 How service quality concept is integrated with process quality improvement?
How it is different from concept of manufacturing process quality?
Service means many different things in different contexts as compared to product quality
(Garvin, 1984). For some it is synonymous with customer care, for others it is equivalent of
logistics function, or internal services such as accounting or personnel, for others it means
10,000 mile check-up to their car.
Despite more than 25 years of study, scholars in the field of services management do not agree
on ‘what a service is’. From customers’ perspective, service is a combination of customer
experience and their perception of outcome of service. An experience at a
park, for example, includes experience of rides, restaurants, emotions of enjoyment and
customer view of value for money at the end of the day. From manufacturing operations
perspective or definition, Figure2-25 shows the transformation process in service operation.
Figure 2-25 ServiceOperation: Service Transformation Process
Figure 2-26 Defining Service from Customer’s Perspective
It is important to note that customers also have to make an input to the service. These
customer inputs include their time and effort plus the financial cost (i.e. the price they pay
for the service) (refer Figure 2-26).
Customer and Staff
Material and Facilities
Equipments and Technology
Service Process
Goods and Services
Customer Time
Customer Effort
Financial Cost for Service
Experience+ Service Product
Value
Emotions
Judgments
Intensions
Service Quality
The term service quality is often used to mean different things. Some operations manager use the
term to mean how customer is treated. This is perhaps more accurately called quality of service,
as opposed to service quality, which can mean outcome and experience. Definitions of service
quality include:
Satisfaction
Sometimes service quality is used to mean same as satisfaction, i.e. perceived (experienced)
service quality.
An impression of the organization and its services
Service quality is more often used as a more enduring construct, whereas satisfaction is situation
and experience-specific. Satisfaction has to be experienced (refer Johnston and Clark, 2008),
whereas customers may have views about organization’s service quality without ever having
experienced the service. Service quality can also be expressed as consumer’s overall impression
of relative inferiority/superiority of organization services. Recent empirical work also suggests
that there is an interactive relationship between satisfaction and service quality, i.e. each can
have a moderating effect on other and on post-purchase intentions (Johnston and Clark, 2008).
Quality delivered
When we talk about service quality from an operation's perspective we usually mean the quality
of the service (may be in multiple stages) we deliver, i.e. does it consistently meet the
specification for that service? This, of course, may be different to how a customer sees the
service (their perceived service quality), and thus there may be a mismatch between a customer's
expectations of a service and their perception of its delivery. This mismatch could be the result
of either a mismatch between expectation and delivery and/or a mismatch between delivery and
perceptions which is a simplified version of the gap model developed by Parasuraman et al.
(1985).
Customers’ Expectations
Organizations need to understand expectations of customers and if appropriate, manage
those expectations (refer Figure 2-27). Indeed it may be appropriate to try to built-in
customer’s expectations in order to keep them at the right level that can be met or just
exceeded by service delivery. This is a key challenge for service operations managers.
Figure 2-27 Expectation Scale
Customers expectation are influenced by various factors as illustrated in Figure 2-28.
Figure 2-28 Factors influencing expectation
Customer’s expectation and service quality factors
Price often has a large influence on expectation. Higher the prices, higher are customer
expectation. Alternative services available also help define and set expectations. Marketing
can have a considerable influence on expectations. Marketing, branding and advertising
campaigns help set expectations.
Word-of-mouth marketing can have a profound effect
on customer expectations. Indeed, in some situations, word-of-mouth may
have a stronger influence than organizational marketing.
Customers' mood and attitude can affect the expectations. Someone in a bad mood or with a
poor attitude to an organization may have heightened expectations; someone less concerned and
more tolerant may have a wider zone of tolerance and thus has a wider range of expectations.
As observed, expectations are dynamic. Customers are continually experiencing many service
situations and their expectations are under continual review and change.
Service quality factors are those attributes of service about which customers may have
expectations and which need to be delivered at some specified level. Several sets of factors have
been identified (Parasuraman et al. 1985, 1988; Zeithaml et al., 1991) while explaining the
service quality gap model. There are 18independent quality factors (Johnston and Clark, 2008)
which try to capture totality of service quality. These factors include access, aesthetics,
attentiveness/helpfulness, availability, care, cleanliness/tidiness, comfort, commitment,
communication, competence, courtesy, flexibility, friendliness, functionality, integrity,
reliability, responsiveness and security.
Module II. Process Quality Improvement
Lecture-9 How six sigma philosophy is aligned with process quality improvement?
High-technology products with many complex components typically have many opportunities
for failure or defects to occur. Motorola developed the six-sigma program in the late 1980s as a
response to the demand for these products. The focus of six-sigma is reducing variability in key
product quality characteristics, or so-called CTQ, to the level at which defects are extremely
unlikely.
Figure 2-29 shows a normal probability distribution as a model for a quality characteristic with
the specification limits at six standard deviations on either side of the mean.
Figure 2-29 Normal probability distribution
Now it turns out that in situation when specification lines are at three standard deviation level,
the probability of producing a product within these specifications is 0.9973, which corresponds
to 2700 parts per million (ppm) defective. This is referred to as three-sigma quality performance.
In case we have a product that consists of an assembly of 100 components or parts and all 100 of
these parts must be no defects for the product to function satisfactorily. The probability that the
unit of product is having no defects is
0.9973 x 0.9973 x ... x 0.9973 = (0.9973)100= 0.7631
That is, about 23.7% of the products produced under three-sigma quality will be defective. This
may not an acceptable situation, because many high-technology products are made up of
thousands of components. An automobile has about 200,000 components and an airplane has
several million.
The Motorola six-sigma concept is to reduce the variability in the process so that specification
limits are six standard deviations from the mean. Then, a
s given in Figure 2-30, there will only be about 2 parts per million defects. For six-sigma
quality, the probability that any specific unit of the hypothetical product is having defect is
0.9999998, or 0.2 defect parts per million.
When the six-sigma concept was initially developed, an assumption was made that when the
process reached the six-sigma quality level, the process mean was still subject to disturbances
that could cause it to shift by as much as 1.5 standard deviations off target. Under this scenario
(shown in Figure 2-30), a six-sigma process would produce about 3.4 ppm defects.
Figure 2-30 Shift in Mean
Table 2-5 PPM level vs sigma rating
Specification limit ±1 σ ±2 σ ±3 σ ±4 σ ±5 σ ±6 σ
Percent inside specs 30.23 69.13 93.32 99.379 99.9767 99.9966
ppm defective 697700 308700 66810 6210 2330 3.4
Considering shift in mean as natural phenomenon, the nonconforming and part per million
(PPM) is given in Table 2-5.
Six sigma also considers an important concept of opportunity to define the PPM level. Motorola
established six-sigma as both an objective for the corporation and as a focal point for process and
product quality improvement efforts. In recent years, six-sigma has spread beyond Motorola and
has come to encompass much more. It has become a program for improving corporate business
performance by both improving quality and paying attention to reducing costs. Companies
involved in a six-sigma effort utilize teams to work on projects that have both quality and
significant financial impact. The effort is better focused than in earlier TQM programs, and has
been more successful in obtaining management commitment. However, remember Deming's
point 10, which essentially says to eliminate slogans and programs to improve quality. There are
many programs including zero defects, value engineering, quality is free, TQM, and so forth,
which has failed due to improper implementation. A major component in successful quality
improvement is driving the use of the proper statistical and engineering tools into the
right places in the organization. A DMAIC (Define, Measure, Analyze, Improve and Control)
approach is used to implement six sigma philosophy. However, it is to be remembered that
statistical approach (say SPC, DOE) is the key theme for variation reduction in Six Sigma
philosophy. Some additional information on Six Sigma and Product Quality Improvement is
provided in Module 3.
Module II. Process Quality Improvement
Lecture – 10 How ISO 9001 and other standards used in Quality Management System
(QMS) influences process quality?
ISO is an international organisation for standardisation, which has been formed for the
development and issuing of international standards to be used across the world. Since its
inception, it has published more than 19,000 standards. Standardization actually helps in the
optimization of operations by proper utilization of resources. Earlier, when ISO started its
operations, it was working as International Federation of the National Standardising Associations
(ISA). But this organisation was dissolved during World War II. The acronym ISO is derived
from the Greek word “isos” which means “equal”. The members of ISO are the recognised
standard authorities, which also represents their respective nations. For example, American
National Standards Institute (ANSI) is the representative of the United States in ISO, and Bureau
of Indian standards is the representative of India. The structure of ISO is comprised of technical
committees, sub-committees and working groups.
The ISO 9000 Series of Standards is generic in scope. By design, the series can be tailored to fit
any organization's needs, whether it is large or small, a manufacturer or a service organization.
ISO 9000 series is developed to serve the quality aspects, which also include the eight principles
of management systems. It can be applied to construction, engineering, health care, legal, and
other professional services as well as the manufacturing of anything from nuts and bolts to
spacecraft. Its purpose is to unify quality terms and definitions used by industrialized nations and
use those terms to demonstrate a supplier's capability of controlling its processes. In very
simplified terms, the standards require an organization to say what it is doing to ensure quality,
then do what it says, and, finally, document or prove that it has done what it said. The main
reason behind establishing ISO standards is to ensure the required safety, quality, and reliability
of products and services. This can raises the levels of productivity and reduce the chance of
error.
The three initial standards of the series are:
ISO 9000:2000-Quality Management Systems (QMS)-fundamentals and vocabulary discusses
the fundamental concepts related to the QMS and provides the terminology used in the other,
two standards.
ISO 9001:2000-Quality Management Systems (QMS)-requirements is the standard
used for registration by demonstrating conformity of the QMS to customers, regulatory, and the
organization's own requirements. ISO 9001:2008 is further developed with the aim of
establishing the requirements of quality management systems. The certification for ISO 9001
needs to be renewed by organization after a particular period as suggested by the certification
body. Generally, this period is three years.
ISO 9004:2000-Quality Management Systems (QMS)-guidelines for performance
improvement provides guidelines that an organization can use to establish a QMS
focused on improving performance. ISO 9004:2009 was created as the latest revision to replace
ISO 9004:2008 and was released in November, 2009.
The standard has eight clauses: Scope, Normative References, Definitions, Quality
Management Systems, Management Responsibility, Resource Management, Product
and/or Service Realization, and Measurement, Analysis, and Improvement. The first
three clauses are for information while the last five are requirements that an organization must
meet.
AS100
This aerospace industry quality system was officially released by the Society of Automotive
Engineers in May 1997. Its development and release represents the first attempt to unify the
requirements of NASA, DOD, and FAA, while satisfying the aerospace industry's business
needs. In March 2001, the International Aerospace Quality Group (IAQG) aligned AS9100 with
ISO 9001:2000.
QS 9000
The famous “Big Three” of US automobile sector, namely Chrysler, Ford and General Motors
had their own supplier development models and associated quality assurance systems initially.
However, in the late 80, a need was felt to develop a harmonized common model for the
suppliers to these big-three as suppliers were faced with the problem of complying to different
quality assurance models for supplies made to different buyers. Accordingly a joint Automobile
Industry Action Group (AIAG) was set up to develop a harmonized supplier development and
quality assurance model, primarily meant for suppliers to the above mentioned automobiles
manufacturers and subsequently also to other automobile related industries throughout the world.
The basis used by the AIAG was the ISO-9000 series of standards in addition to the existing
individual quality assurance standards of Chrysler, Ford and General Motors. These are
Chrysler : Supplier Quality Assurance Manual
Ford : Q-101 Quality System Standard
General Motors : North American Operations Targets for Excellence
QS-9000 (http://en.wikipedia.org/wiki/QS9000) defines the fundamental quality system
expectations of Chrysler, Ford, General Motors, Truck Manufactures and other subscribing
companies for internal and external suppliers of production and service parts and materials.
These organizations are committed to working with suppliers to ensure customer satisfaction
beginning with conformance to quality requirements, and continuing with reduction of variation
and waste to benefit the final customer, the supply base, and themselves.
In addition, ISO 90003:2004 is another standard that is developed with the aim of improving the
quality of software-related products in terms of supply, development, maintenance and support
services. Whereas, ISO 13485 standards published states all the specifications required for a
comprehensive quality management system that helps in the design and manufacturing of
medical devices. As quality practices also influences society at large, ISO 14000 was created
with the aim of controlling the adverse effects to environment occurring due to the processes
followed by organizations. ISO 14001 standards are designed as the representation of all the
standards that are used for the successful implementation of Environmental Management
System.
All the standards for quality management discussed above talks about
process/system/environment improvement as the central theme and provide a general guideline
to achieve the goal.
Module II. Process Quality Improvement
Lecture – 11 What is the role of Audit in implementation of Quality Management System
(QMS)?
A QMS encompasses a whole set of activities, which need to be coordinated and controlled for
producing the desired quality in an organization. The objective of QMS is to improve the
performance of an organization and sustained improvements. In this context, internal audits
allow for checking the effectiveness of the installed quality systems from time to time. Internal
audits help in ensuring that the QMS conforms to the organizational quality policies as well as
any quality standard that it follows. Management review are also adapted by organization to
check the current capacity of QMS being implemented. These reviews may be conducted after
four to six months of QMS implementation. The review also helps to keep the track of adequacy
of the QMS. Before applying for ISO 9000 or any other international standard certification, a
pre-assessment audit is also done at organization level by external agency. Pre-assessment audit
helps the organisation understand where they stand with respect to quality requirements, and
provide an opportunity to correct issues which may come in the way of a successful international
standard certification application. Any kind of quality audit helps to maintain ‘what we record
and what we actually do’.
Module II. Process Quality Improvement
Lecture – 12 How Quality Awards, Quality council, Quality circle, Quality
Improvement Teams helps improve process quality?
Quality awards are prizes awarded for some aspect of quality performance that has been
demonstrated by an organization. Deming Prize was instituted by the Union of Japanese
Scientists and Engineers (JUSE) in 1951 to honour the contributions of W.E. Deming
towards quality control in Japan. This award is given to organization who adopted and
preached Total Quality Management principals successfully. Simililar in line, Malcolm
Baldrige National Quality Award (MBNQA)was instituted in USA in the year 1987. In India,
Bureau of Indian Standards constituted Rajiv Gandhi National Quality Award, to promote
excellence in Indian manufacturing and service organisations,in 1991.
There are seven parameters for evaluating any organisation for qualifing for any specific
award and marks are allocated for each one of the parameter and its sub-parameters. Any of
the quality award emphasis on strategic planning and links between strategic and quality
planning. Overall emphasis is on improving processes and system for better quality output.
In 1960, quality circles are formed in Japanese industries with the aim of improving the levels
of quality. This concept is still followed in many manufacturing and service industrial for
quality and process improvement. A quality circle is basically a participatory management
philosophy, in which a group of employees is formed to identify, analyse and solve quality
problems.The main objectives of quality circle is to identify quality issues, find out the root
causes, and solve the issues to improve process quality.
1
Module III Product Quality Improvement
Lecture – 1 How QFD helps in product quality improvement?
Quality Function Deployment (QFD) or the house of quality is the foundation to link the voice of
the customers with technical design requirements of a product. In other words, abstract
specifications required by the targeted customers are translated into specific product technical
requirements. Say in summer, customer needs a room to be cool and comfortable. However, how
much cool gives comfort to him/her is not specified. Take another situation, in which, a customer
wants hot coffee. Hot coffee is one of the ‘voice of the customer (VOC)’ [or ‘critical-to-quality
(CTQ) characteristic’] that the customer demands. He /She may not specify the temperature, but
the shopkeeper needs to identify best possible temperature setting for the coffeemaker machine.
The best setting will also differ according to weather conditions/ seasons. In order to translate a
VOC (say, comfort temperature range for AC), the AC machine designer must first experiment
and specify the feasible range of temperature setting (say 180C to 270C) for varied customers.
Providing varied temperature setting leads to flexibility in the design and helps different
customer to set different comfortable temperature at workplace/ home. There can be more than
one VOC, which can also be interacting. So, as the understanding on customer’s priorities /needs
(VOC) for a product becomes clearer and subsequently freezed, the designer attempts to translate
those into product technical requirements, so as to deliver the best tradeoff solution for
interacting VOC. The next test is to build a product prototype and check real life performance of
the machine. This is a continual design improvement process activity and finalizing a design may
require 30 to 40 prototype experimentation. Subsequently, the product design is approved for
pilot/full production. QFD is a structured framework to translate the VOC to technical
specification of a product. It is not an optimization tool, and does not provide any tradeoff
solution. It only guides the engineers towards developing a robust product design from the
customer’s perspective.
The structure of QFD can be thought of as a house (so-called ‘House of Quality’), and shown in
Figure 3-1.
Figure 3-1: House of Quality
The parts of the house of quality are described as:
The outside walls of the house are shown as the customer requirements and their priorities.
On the left side is a listing of VOC. On the right side is the prioritized customer requirement,
which is derived from customer survey. The ceilings of the house contain the technical
descriptors or requirements with expert’s priorities. The central or interior walls of the house
are the relationships between customer requirements and technical requirements. Customer
voices (customer requirements) are translated into engineering requirements (technical
descriptors).
The roof of the house is the interrelationship between independent technical requirements. Here
the trade-offs between similar and/or conflicting technical requirements are identified. The aim
of the house is to determine prioritized technical requirement. Technical benchmarking, reverse
engineering, tradeoff, and target value comparison are mostly used to determine technical
bounds.
This is the basic structure for the house of quality. However, based on this format varied QFD
matrices are proposed.
Building a House of Quality
Quality function deployment starts with a list of goals/objectives. This list is often referred as the
WHATs that a customer needs or expects in a particular product. This list of primary customer
requirements is usually vague and very general in nature. Further definition is accomplished by
defining a new, more detailed list of secondary customer requirements needed to support the
primary customer requirements. In other words, a primary customer requirement may encompass
numerous secondary customer requirements.
Let us consider the development process of designing a handlebar stem for a bicycle.
Let us assume that there are two primary customer requirements, viz. aesthetics and
performance. The secondary customer requirements under aesthetics are affordable cost,
aerodynamic look, proper finish, and corrosion resistance. The secondary customer requirements
under performance are light weight, strength, and durability. This is illustrated in the QFD or
House of Quality diagram (Figure 3-2).
Figure 3-2 House of Quality of a handlebar stem in a bicycle
As the customer needs and expectations are expressed in terms of customer requirements, the
QFD team needs to come up with engineering characteristics (HOW’s)
that will affect one or more of the customer requirements. Each engineering characteristic must
directly affect a customer perception (VOC) and be expressed in measurable terms.
Implementation of the customer requirements in design is difficult until they are translated into
counterpart technical characteristics. Counterpart technical characteristics are an expression of
the voice of the customer in technical language and specifications. For example, a customer
requirement for an automobile might be a smooth ride. This is rather an abstract statement,
which is important from the point of view of selling an automobile. Technical characteristics for
a smooth ride can be appropriate dampening, anti-roll, and stability requirements. These are the
primary technical descriptors or characteristics. Engineering knowledge and brainstorming
among engineering staff’s is a suggested method for determining technical characteristics.
Figure 3-3 shows the different technical requirements which can address all VOC for the bike
stem design.
Figure 3-3 Interrelationship between VOC and Technical Requirements
The next step in building a house of quality is to compare the VOC with technical characteristics
and determine their interrelationships. In this context, engineering knowledge about the product
and historic evidence/ data can provide useful information. Common practice is to use symbols
to represent the nature of relationship between customer requirements and technical descriptors.
Symbols used are:
I. A solid circle represents a strong relationship (scored as +9).
II. A single circle represents a medium relationship. (scored as +3).
III. A triangle represents a weak relationship (scored as +1).
IV. The box is left blank if there is no relationship between VOC and technical
characteristics.
Figure 3-4 provides the interrelationship matrix with type of relationships. Any cell that is
empty implies no or insignificant relationship.
Figure 3-4 Complete Interrelationship between VOC and Technical Requirements
After drafting the relationship matrix, it is evaluated for any empty row or column. An empty
row indicates that a customer voice is not being addressed by any technical descriptors. Thus, the
customer expectation is not being met. Any blank column indicates that the technical
requirement is unnecessary, as it does not address any VOC.
The roof of the house of quality, expressed as correlation matrix, is used to identify any
interrelationships between the technical descriptors (Figure 3-5). Symbols are used to describe
the strength of the interrelationships. Symbols generally preferred are:
I. A ‘solid circle’ represents a strong positive relationship.
II. A ‘circle’ represents a positive relationship.
III. An ‘X’ represents a negative relationship.
IV. An ‘asterisk’ represents a strong negative relationship.
Figure 3-5 Correlation Matrix and Tradeoff between Technical Requirements
The symbols also describe the direction of the correlation. In other words, a strong positive
interrelationship means nearly perfect positive correlation. A strong negative will indicate nearly
perfectly negative correlation. This type of representation allows the user to identify which
technical characteristics support one another and which are conflicting. Conflicting technical
descriptors are extremely important because they are frequently the result of conflicting customer
requirements and, consequently, represent points at which tradeoffs must be made. Tradeoffs that
are not identified and resolved, while defining specification, will often lead to unfulfilled
requirements, unnecessary engineering changes, increase in cost, and poor quality from the
standpoint of customers. Some of the tradeoffs may require high-level managerial interventions,
because they cross functional boundaries.
An example of tradeoffs in the design of a car is customer requirements of
high fuel economy and safety. These two CTQ and technical descriptors are conflicting.
Addition of stronger bumpers, air bags, and antilock brakes will ultimately reduce the fuel
efficiency of the car.
The customer’s competitive assessment (Figure 3-6) is a pair of table (or graph) that depicts how
competitive products compare with current organization product status on specific VOC. The
customer competitive assessment is the block of columns corresponding to each
customer requirement in the house of quality on the right side of the relationship matrix,
The numbers 1 through 5 are listed in the competitive evaluation column to indicate a rating of 1
for worst and 5 for best. The customer competitive assessment is a good way to determine if the
customer voice has been met (as compared to best competitor) and identify areas to improvement
for future design.
Figure 3-6 Competative Assessment of VOC
The technical competitive assessment makes up a block of rows corresponding to each technical
descriptor in the house of quality beneath the relationship matrix. After respective technical
factors have been established, the products are evaluated for each technical factor that addresses
VOC.
Similar to the customer competitive assessment, the data recorded are in a scale of 1 through 5,
to indicate a rating, 1 for worst and 5 for best. The technical competitive assessment is often
useful in uncovering gaps in engineering judgment.
Importance ratings represent the relative importance of customer requirement in
terms of each other.
The target-value of column can be on the same scale as the customer competitive assessment (1
for worst, 5 for best can be used). This column is where the QFD team decides whether they
want to keep their product unchanged, improve the product, or make the product better than the
competitor.
The prioritized technical descriptors make up a block of rows corresponding to the technical
descriptor in the house of quality below the technical competitive assessment as shown in Figure
3-7. These prioritized technical descriptors contain target value and absolute weights.
Figure 3-7 Absolute Weights of Technical Requirements
The last rows of the prioritized technical descriptors are the absolute weight. A popular and easy
method for determining the weights is to assign numerical values to symbols in the relationship
matrix symbols. The absolute weight for the jth technical descriptor is given as
∑=
=n
iiijj cRa
1
Where,
aj = row vector of absolute weights for the degree of technical difficulty of technical
descriptors
(i = 1, ... , m)
Rij = weights assigned to the relationship matrix (i = 1, ... , n, j = 1, ... , m)
ci = column vector of importance to customer for the customer requirements
(i = 1, ... , n)
m = number of technical descriptors
n = number of customer requirements
The absolute weight for each technical descriptor is determined by taking the dot
product of the column in the relationship matrix and the column for importance to customer. For
instance, for aluminum (see Figure 3-7) the absolute weight is
(9x8+1x5+9x5+9x2+9x7+3x5+3x3) x1 =227.
The greater values of absolute weight indicate higher importance of the technical descriptor to
address VOC. These weights can be organized into a Pareto diagram to show which technical
characteristics are most important in meeting customer requirements.
In a corrosion problem, a Japanese car company Toyota, during 1960’s and 1970, there was huge
expense on warranty. The Toyota Rust QFD Study resulted in a virtual elimination of corrosion
warranty expenses. The customer requirement on durability was also achieved, with no visible
rust in following three years. It was determined that this could be obtained by including a
minimum paint film build, and maximum surface-treatment. The key process operation that
provides these part-quality characteristics consists of a three-coat process.
Module III Product Quality Improvement
Lecture – 2 What are component and system reliability and how it can be improved?
Reliability is a measure of the quality of the product over the long run. The concept of reliability
is an extended time period over which the expected operation of the product is considered and
we expect the product will function according to certain expectations over a stipulated period of
time. With the customer and warranty costs in mind, we must know the chances of successful
operation of the product for at least a certain stipulated period of time. Such information helps
the manufacturer to select the parameters of a warranty policy.
Technically, reliability is the probability of a product performing its intended function for a
stated period of time under certain specified conditions. Four aspects of reliability are apparent
from this definition first, reliability is a probability of success-related concept; the numerical
value of this probability is always between 0 and 1. Second, the functional performance of the
product had to be measured under certain stipulated conditions. Product design is expected to
ensure development of a product that meets or exceeds the specified requirements under
specified operating conditions. For example, if the breaking strength of a nylon cord is expected
to be 1000 kg, then in the predefined operational conditions, the cord must be able to bear
weights of 1000 kg or more. Third, reliability implies successful operation over a certain period
of time (t). Although no product is expected to last forever, the time dimension ensures
satisfactory performance over at least a minimal stated period (say, 100 hours). In the context of
these three aspects, the reliability of the nylon cord might be described as having a probability of
successful performance of 0.92 in bearing loads of 1000 kg for 1 year under dry conditions.
It is observed that most manufacturing products go through three distinct phases (see Figure 3-
8) from product inception to wear-out.
Figure 3-8 Bathtub Curve
The life-cycle curve of Figure 3-8 shows the variation in the failure rate as a
function of time in different phases. Conventionally the failure rate (λ ) is plotted as a function
of time. This curve is often referred to as the bathtub curve; it consists of the debugging (infant-
mortality) phase, the chance-failure phase (useful life phase), and the wear-out phase.
The debugging phase, also known as the infant-mortality phase, exhibits a drop
in the failure rate as initial problems identified during prototype testing are removed.
The chance-failure phase, between times t1 and t2, is then encountered; failures occur
randomly and independently. This phase, in which the failure rate is constant, typically
represents the useful life of the product delivered to end customer. In the wear-out phase, an
increase in the failure rate is expected due to wear and tear of the product. Here, after the end of
their useful life, parts age and wear out.
For the random chance-failure phase, which represents the useful life of the
product or component, the failure rate is assumed to be constant. As a result, the exponential
distribution is selected to describe the time-to-failure of the product for this phase.
An exponential distribution as a memory less property and its probability density function is
given by
( ) λλ −= ≥, t 0tf t e
The mean-time-to-failure (MTTF) for the exponential distribution can be expressed as
MTTF=1/λ
The reliability, at time t, say R(t), is the probability of the product lasting up to time t. It can be
expressed as,
( ) ( )t
0
1
=1- t t
R t F t
e dt eλ λ− −
= −
=∫
Here, F(t) represents the cumulative distribution function at any time t. Reliability decreases
exponentially with time (Figure 3-9) and the failure-rate function, say r(t) , is given by the ratio
of the time-to-failure probability density function to the reliability function. We have
( ) ( )( )
f tr t
R t=
Figure 3-9 Reliability v/s Time
Thus, assuming an exponential distribution, r (t) implying a constant failure rate, as
shown below.
( )λ
λ
λ λ−
−= =
t
t
er te
Let us consider a resistor component, which follows an exponential time-to-failure distribution
with a failure rate of 8% per 1000 hr. We are interested to calculate the reliability of the resister
at 5000 hr, and also we intend to calculate the mean-time-to-failure. Here the constant failure
rate λ is obtained as
λ = 008 /1000 hr =0.00008/hr
Thus, the reliability for 5000 hr of survival is
( )( )( )- 0.00008 5000
-0.4
=e =e 0.6703
tR t e λ−=
=
Thus there is about 67% chance of survival (success) of the resister under stipulated conditions
and stipulated time (5000 hr). The mean (average) time-to-failure (assuming it cannot be repair)
of the resister will be
MTTF=1/ 1 /0.00008 12,500 hλ = =
System Reliability
Let us consider a system with three components (say three resister) in series as shown in Figure
3-10.
Figure 3-10 Series System
Without loss of generality, if the system components can be assumed to have a time-to-failure
distribution as exponential with each component has a constant failure rate, we can easily
compute the reliability of n-system in series. Suppose the system has n components and in series,
each with exponentially distributed time-to-failure with failure rates 1 2, , , nλ λ λ . The system
reliability is calculated as the product of the component reliabilities:
λ λ λ
λ
− − −
=
= × × ×
∑
1 2
1
( )
=exp -
nt t tS
n
ii
R t e e e
t
This implies that the time-to-failure of the system is exponentially
distributed with an equivalent failure rate of 1
niiλ
=∑ . The mean time to failure for the system is
given by
1
1n
ii
MTTFλ
=
=
∑
Systems with Components in Parallel
System reliability can be improved by placing redundant components in parallel. The system
operates as long as at least one of the components operates. A three component parallel system is
represented as given in Figure 3-11.
Figure 3-11 A Parallel Component System
If the time-to-failure of each component follows exponential distributions, each with a constant
failure rate, iλ , i = 1, ... , n, the system reliability, assuming independence of component
operation (failure of one does not impact failure of any other component), is given by
( )
( )λ−
= −
−
∏
∏
n
i=1n
i=1
( ) 1- 1 ( )
=1- 1 i
S i
t
R t R t
e
In a special case, where all components have the same failure rate, λ , the system
reliability for parallel component is given by
( )λ−= − −( ) 1 1 int
SR t e
For such specific situation, the mean-time-to-failure for the system with n identical components
in parallel, and also assuming that each failed component is immediately replaced by an identical
component, can be expressed as
1 1 1 11 ...2 3
MTTFnλ
= + + + +
Systems with Components in Series and in Parallel
Real life systems often consist of components that are mixed and consist of both series and
parallel configuration. For such system, reliability calculation is primarily based on the
previously discussed concepts, and assumption of components operating independently. Parallel
systems are first collated to get a composite reliability, and then the overall components are
considered as series to calculate system reliability. Systems can also consist of standby
component, which operates as and when base component fails. Reader may refer to book by
Amitava Mitra (2008) or Besterfield et al (2004 ) for further details on. K-out-of-N system
(parallel system is 1 out of N system) is another possible system configuration.
Module III Product Quality Improvement
Lecture 3-What is Design FMEA?
Continually measuring the reliability of a product is an essential
part of Quality. When creating a new product, or even modifying an existing product, it is
always necessary to improve the reliability of the product. One of the most powerful methods
available for improving the reliability of product is design FMEA. FMEA is an approach that
combines the technology and experience of people in identifying foreseeable failure modes of a
product and planning for its elimination. FMEA attempts to detect the potential product-related
failure modes. The approach is used to anticipate causes of failure and prevent them from
happening. It is like eliminating/preventing potential causes of failure in a cause and effect
diagram. This method can be implemented in both the product design and process design and
involves effect on both internal and the external customer.
FMEA uses an occurrence and detection probability criteria in conjunction with severity criteria
to develop a risk prioritization numbers for prioritization the corrective action. It is to be noted
that for FMEA to be successful, it is extremely important to treat the FMEA as a living record,
and continually changing as per new problem(s) and being updated to ensure that the most
critical problems are identified and addressed to prevent recurring.
A design (product) FMEA or process FMEA can provide the following benefits:
(i)Having a systematic review approach of component failure modes can ensure that any failure
produces minimal damage to the product or process.
(ii) Determining the effects that any failure will have on product or process and their functions.
(iii) Determining those critical parts of a product or a process whose failure will have critical
effects on product or process operation.
(iv) Eliminating or minimizing the adverse effects that failures could generate and indicating
safeguards to be incorporated if the product or the process cannot be made fail-safe or brought
within acceptable failure limits.
(v) Help uncover oversights, misjudgments, and errors that may have been made.
It is to be noted that a FMEA document, however, cannot solve all design and process problems
and failures. The document, by itself, will not fix the identified problems or define the action
that needs to be taken. FMEA cannot also replace the basic root cause analysis approach.
FMEA Team
The FMEA approach is a team effort where the responsible engineer involves design,
manufacturing, materials, quality, service, supplier, and even the next customer (whether
internal or external). The team leader has certain responsibilities, which include coordinating
corrective action assignments and follow-up, keeping files and records of FMEA forms, leading
the team through completion of the forms, keeping the process moving, and finally, drawing
everyone into participation.
Details on FMEA Documentation
The concept of FMEA is nothing new to engineers. Engineers designing and building a product
have always incorporated the concepts of FMEA in their thinking process. However, FMEA
does help keep those ideas available for future use and for the use of others. One engineer may
find a potential problem elementary and not worth extra attention; a second engineer may not
realize the problem altogether. The purpose of the FMEA document (Please see Figure 3-12) is
to allow all involved engineers to have access to others' thoughts and to design and manufacture
using this collective group of thoughts. In this document, on the top right corner (see Figure 3-
12) is the FMEA Number. This number is only for record. There is also an item space to clarify
which exact component or process is being analyzed. The name and number of the system or
sub-system being analyzed is also mentioned in this space. Some of the critical headings
mentioned in FMEA document is discussed below.
Figure 3-12 Design FMEA Document
Design Responsibility
The team in charge of the design or process is identified in the space designated as Design
Responsibility. The name and department of the person or group responsible for preparing the
documentation is included here.
Prepared By
The name, telephone number, and address of the concerned persons (group) are included here so
as to contact them in case a part of the document needs further explanation.
FMEA Date
The date the FMEA was compiled and the latest revision date is included in this FMEA Date
space.
Item/Function
In this section, the name and part number of the item being analyzed is recorded. This
information avoids confusion involving similar items. Next, the function of the item is to be
entered below the description of the item. No specifics should be left out in giving the function
of the item. If the item has more than one function, they should be listed here. The function of
the item including the environment in which the system operates (say temperature, pressure, and
humidity) is also recorded here.
Potential Failure Mode
The Potential Failure Mode information may be one of two things. First, it may be the way in
which the item may fail to meet the design criteria. Second, it may be a potential failure in a
higher-level system or may be the result of failure of a lower-level system. It is important to
consider and list each and every potential failure mode. A possible starting point when listing
potential failure modes is to consider past failures. Also, the potential failure modes must be
described in technical terms. Some typical failure modes may include ‘cracked or deformed,
loosened joints, leakage from welding, short circuit in water heater, and fractured.'
Potential Effect(s) of Failure
The potential effects of failure are the effects of the failure as perceived by the internal or
external customer. The effects of failure must be described in terms of what the customer will
notice or experience. It is also stated whether the failure will impact personal safety or violate
any product regulations. This section of the document must also forecast what effects the
particular failure may have on other subsystems in immediate contact. Some typical effects of
failure may include engine noise and poor appearance.
Severity (S)
Severity is the assessment of the seriousness of the effect of the potential failure mode
to the subsequent component, sub-system, or customer. It is to be emphasized that the severity
applies only to the effect of the failure, not the potential failure mode. Severity rating must not
change from any reasoning except change in the product design. Severity is rated on a 1-to-10
scale, with a 1 being least severe and a 10 being the most severe. Rating criteria is given in
Table 3-1. Readers may also refer QS 9000 (http://en.wikipedia.org/wiki/QS9000), which
provides further details on severity rating.
Classification (Class)
This column is used to classify any special characteristics for components that may require
additional controls.
Potential Cause(s)/Mechanism(s) of Failure
Every potential failure cause is to be listed completely and concisely. Some failure modes may
have more than one cause and/or mechanism of failure. Typical failure causes may include
incorrect product specification, inadequate design, over-stress, poor environment protection.
Typical failure mechanisms may be creep, fatigue, wear, and corrosion.
Occurrence (0)
Occurrence is the possible chance that one of the specific causes/mechanisms will occur. This is
done for every cause and mechanism listed. Reduction or removal in occurrence ranking must
not come from any reasoning except for a direct change in the design or process. Change is the
only way a reduction in the occurrence ranking can be affected. The likelihood of occurrence is
based on a 1-to-10 scale, with 1 being the least chance of occurrence and 10 being the highest
chance of occurrence. A reference on occurrence rating is given in Table 3-2.
Table 3-1 Severity Rating Reference
Effect Criteria: Severity of Effect Ranking
Hazardous Without warning
Very high ranking when potential failure mode affects safe operation and/or regulation noncompliance. Failure occurs without warning.
10
Hazardous With warning
Very high ranking when potential failure mode affects safe operation and/or regulation noncompliance. Failure occurs with warning.
9
Very High Item or product is inoperable, with loss of function. Customer very dissatisfied.
8
High Item or product is operable, but with loss of performance. Customer dissatisfied.
7
Moderate Item or product is operable, but with loss to comfort/convenience items inoperable. Customer experiences discomfort.
6
Low Item or product is operable, but with loss of performance of comfort/convenience items. Customer has some dissatisfaction.
5
Very Low Certain item characteristics do not conform. Noticed by most customers.
4
Minor Certain item characteristics do not conform. Noticed by average customer.
3
Very Minor Certain item characteristics do not conform. Noticed by discriminating customers.
2
None No Effect. 1
Table 3-2 Occurance Rating Reference
Probability of Failure Possible Failure Rates Ranking
Very High: Failure is
Almost Inevitable.
>1 in 2 10
1 in 3 9
High: Repeated Failures
1 in 8 8
1 in 20 7
1 in 80 6
1 in 400 5
1 in 2000 4
Low: Relatively Few
Failures
1 in 15,000 3
1 in 150,000 2
Remote: Failure is
Unlikely <1 in 1,500,000 1
Current Design Controls
In order to improve the occurrence rating for the particular failure mode, the design control must
be employed. Current Design control indicates the state of control that will be able to detect the
occurance of a failure or minimize the failure chances.
Detection (D)
This is a relative measure of assessment of the ability of the design control to detect either a
potential cause/mechanism or the subsequent failure mode before the component goes to
end/next user. Typically, in order to achieve a lower detection rating, design control must be
improved. A reference for rating in detection phase is given in Table 3-3.
Table 3-3 Rating of Likelihood of Detection in Design FMEA
Rankings of likelihood of detection by Design Control for Design FMEA
Effect Criteria : severity of Effect Ranking Absolutely Impossible
Design control will not and / or cannot detected a potential cause / mechanism and subsequent failure mode : or there is no design control
10
Very remote
Very remote chance the design control will detected a potential cause /mechanism subsequent failure mode
9
Remote Remote chance the design control will detect a potential cause / mechanism and subsequent failure mode.
8
Very low Very low chance the design control will detect a potential cause /mechanism and subsequent and failure mode
7
Low Low chance the design control will detected a potential cause / mechanism and subsequent failure mode
6
Moderate Moderate chance the design control will detect a potential cause / mechanism and subsequent failure mode
5
Moderate highly
Moderately high chance the design control will detect a potential cause / mechanism and subsequent failure mode
4
High High chance the design control will detect a potential cause / mechanism and subsequent failure mode
3
Very High Very high chance the design control will detect a potential cause /mechanism and subsequent failure mode
2
Almost certain
Design control will almost certainly detect a potential cause / mechanism and subsequent failure mode
1
Risk Priority Number (RPN)
The Risk Priority Number is the product of the severity (S), occurrence (0), and detection (P)
rankings. This product may be viewed as a relative measure of the design risk. Values for the
RPN can range from 1 to 1000, with 1 being the smallest design risk possible. This value is then
used to rank order the various causes of failure in the design. For causes with a relatively high
RPN, the engineering team must make efforts to take corrective action to reduce the RPN. Any
score above 50 may be considered as cutoff to eliminate/minimize the impact of a particular
cause. However, because a certain concern has a relatively low RPN (<50), the FMEA team
should not overlook the concern and neglect an effort to reduce the RPN. This is especially true
when the severity of a concern is high. In such case(s), a low RPN may be extremely misleading,
not placing enough importance on a concern where the level of severity may be disastrous. In
general, the purpose of the RPN is to rank the various causes on the record. However, every
cause should be given full priority by the team, and the team should look for every method
available to reduce the RPN.
Recommended Actions
After every concern has been examined and given a risk priority number, the team should begin
to examine the corrective action(s) that may be employed, beginning with the concern with the
greatest RPN and working in descending order according to RPN. Also, concerns with high
severity should be examined. The purpose of the recommended actions is to reduce one or more
of the rating that constitute the risk priority number. An increase in design validation actions will
result in a reduction in only the detection ranking. Only removing or controlling one or more of
the causes/mechanisms of the failure mode through design revision can effect a reduction in the
occurrence ranking. And only a design revision can bring about a reduction in the severity
ranking. Some actions that should be considered when attempting to reduce the three rankings
include, but are not limited to: design of experiment (DOE), revised test plan, and revised
design.
Responsibility and Target Completion Dates
Here the individual or group responsible for the recommended actions and the target completion
date should be entered as reference for future record.
Actions Taken
After a corrective action has been implemented, a brief description of the action and its effective
date is entered. This is done after the action has been implemented so future users can track the
progress of the plan.
Resulting RPN
After the corrective actions have been identified, the resulting severity, occurrence, and
detection rankings should be re-estimated. Then the resulting RPN should be reo calculated and
recorded. If no actions are taken, this section should be left blank. If no actions are taken and the
prior rankings and RPN are simply repeated, future users may reason that there were
recommended actions taken, but that they had no effect. After this section is completed, the
resulting RPNs should be evaluated, and if further action is deemed necessary, steps from the
recommended actions section can be repeated.
The overall objective of Design FMEA is to improve the design, improve product reliability, and
reduce the chances of occurrence of failures. Design of Experiment (DOE) is recommended by
various researchers to improve the quality of design. One of the DOE approach is so-called
‘Robust Design’, originally proposed by Genechi Taguchi in 1980. A bried detail on his concept
is discussed below.
Module III Product Quality Improvement
Lecture – 4 What is robust design?
Dr. Genichi Taguchi, a mechanical engineer, who has won four times Deming Awards,
introduced the loss function concept, which combines cost, target, and variation into one metric.
He developed the concept of robustness in design, which means that noise variables (or nuisance
variables or variables which are uneconomical to control) are taken into account to ensure proper
functioning of the system functions. He emphasized on developing design in presence of noise
rather than eliminating noise.
Loss Function
Taguchi defined quality as a loss imparted to society from the time a product is shipped to
customer. Societal losses include failure to meet customer requirements, failure to meet ideal
performance, and its harmful side effects.
Assuming the target [tau (τ )] is correct, losses are those caused by a product's critical
performance characteristics, if it deviates from the target.The importance of concentrating on
"hitting the target" is shown by Sony TV sells example. In spite of the fact that the design and
specifications were identical, U.S. customers preferred the color density of shipped TV sets
produced by Sony-Japan over those produced by Sony-USA. Investigation of this situation
revealed that the frequency distributions were markedly different, as shown in Figure 3-13. Even
though Sony-Japan had 0.3% outside the specifications, the distribution was normal and centered
on the target with minimum variability as compared to Sony-USA. The distribution of the Sony-
USA was uniform between the specifications with no values outside specifications. It was clear
that customers perceived quality as meeting the target (Sony-Japan) rather than just meeting the
specifications (USA). Ford Motor also had a similar experience with their transmissions.
Figure 3-13 Distribution of color density for Sony-USA and Sony-Japan Out of specification is the common measure of quality loss in Goal post mentality [Figure 3-14
(a) ]. Although this concept may be appropriate for accounting, it is a poor concept for various
other areas. It implies that all products that meet specifications are good, whereas those that do
not are bad. From the customer's point of view, the product that barely meets the specification is
as good (or bad) as the product that is barely just out-of-specification. Thus, it appears that wrong
measuring system for quality loss is being used. The Taguchi’s loss function [Figure 3-14 (b)]
corrects for the deficiency described above by combining cost, target, and variation into one
single metric.
Figure 3-14(a): Discontinuos Loss Function (Goal Post
Mentality)
Figure 3-14(b): Continuous Quadratic Loss function (Taguchi Method)
Figure 3-14(a) shows the loss function that describes the Sony-USA situation as per ‘Goal Post
Mentality’ considering NTB (Nominal-the-Best)-type of quality charecteristic. Few performance
characteristics considered as NTB are color density, voltage, bore dimensions, surface finish. In
NTB, a target (nominal dimension) is specified with a upper and lower specification, say
diameter of a engine cylinder liner bore. Thus, when the value for the performance characteristic,
y, is within specifications the quality loss is $0, and when it is outside the specifications the loss
is $A. The quadratic loss function as shown in Figure 3-14(b) describes the Taguchi method of
definining loss function. In this situation, loss occurs as soon as the performance characteristic, y,
departs from the target, τ .
The quadratic loss function is described by the equation
( )2L k y τ= − , Where L = cost incurred as quality deviates from the target (τ )
y is the performance characteristic, k = quality loss coefficient.
The loss coefficient is determined by setting
( )2 2/ /k A y Aτ= − = ∆
Assuming, the specifications (NTB) is10 ± 3 for a particular quality characteristic and the
average repair cost is $230, the loss coefficient is calculated as,
2 2/ 230 /3 25.6k A= ∆ = =
Thus, L = 25.6 (for y= 10) and at L=102.4 (for y = 12),
( )( )
2
2
25.6 10
=25.6 12 10 =$102.40
L y= −
−
Average or Expected Loss
The loss described above assumes that the quality characteristic is static. In reality, one
cannot always hit the target. It will vary due to presence of noise, and the loss function must
reflect the variation of many pieces rather than just single piece. An equation can be derived by
summing the individual loss values and dividing by their number to give
( )22L k yσ τ = + −
Where L = the average or expected loss, σ is the process variability of y charecteristic, y is the
average dimension coming out of the process.
Because the population standard deviation, σ , is unknown, the sample standard deviation, s, is
used as a substitute. This action will make the variability value somewhat larger. However, the
average loss (Figure 3-15) is quite conservative in nature.
Figure 3-15 Average or Expected Loss
The loss can be lowered by reducing the variation, and adjusting the average, y, to bring it on
target.
Lets compute the average loss for a process that produces shafts. The target value , say 6.40 mm
and the loss coefficient is 9500. Eight samples give reading of 6.36, 6.40, 6.38, 6.39, 6.43, 6.39,
6.46, and 6.42. Thus,
s = 0.0315945 6.40375y =
( )
( )
22
22 =9500 0.0315945 6.40375 6.40
=$9.62
L k s y τ = + − + −
There are two other loss functions that are quite common, smaller-the-better and larger- the-
better. In smaller-the-better type, the lesser the value is preferred for the characteristic of interest,
say defect rate, expected cost, and engine oil consumption. Figure 3-16 illustrates the concept.
Figure 3-16 : (a) Smaller –the –Better and (b) Larger–the- Better-type of Loss Function
To summarize the equations for the three common loss functions, Nominal the best
( )2L k y τ= − Where 2/k A= ∆ L =k (MSD) where MSD= ( )2 /y nτ Σ + −
( )22L k yσ τ = + −
Smaller the better 2L ky= where 2/k A y=
L =k(MSD) where MSD= 2 /y n Σ
2 2L k y σ = +
Larger the better
( )21L k y= − where 2k Ay=
L =k(MSD) where MSD= ( )21 / /y n Σ
( )21 / /L k y n = Σ
In case of larget-the-better, higher value is preferred for the characteristic of interest. Few
examples of performance characteristics considered as larget-the-better are bond strength of
adhesives, welding strength, tensile strength, expected profit.
Orthogonal Arrays
Taguchis method emphasized on highly fractionated factorial design matrix or Orthogonal arrays
(OA) [http://en.wikipedia.org/wiki/Orthogonal_array] for experiment. This arrays are developed
by Sir R. A. Fischer and with the help of Prof C R Rao (http://en.wikipedia.org/wiki/C._R._Rao)
of Indian Statistical Institute, Kolkata. A L8 orthogonal array is shown below. An orthogonal
array is a type of experiment where the columns for the independent variables are “orthogonal”
or “independent” to one another.
Table 3-4 L8 Orthogonal Array
The 8 in the designation OA8 (Table 3-4) represents the number of experimental rows, which is
also the number of treatment conditions (TC). Across the top of the orthogonal array is the
maximum number of factors that can be assigned, which in this case is seven. The levels are
designated by 1 and 2. If more levels occur in the array, then 3, 4, 5, and so forth, are used. Other
schemes such as -1, 0, and +1 can be used. The orthogonal property of an OA is not
compromised by changing the rows or the columns. Orthogonal arrays can also handle dummy
factors and can be accordingly modified. With the help of OA the number of trial or experiments
can be drastically reduced.
To determine the appropriate orthogonal array, we can use the following procedure,
Step-1 Define the number of factors and their levels.
Step-2 Consider any suspected interactions (if required).
Step-3 Determine the necessary degrees of freedom.
Step-4 Select an orthogonal array.
To understand the required degree of freedom, let us we consider four two-level (leveled as 1
and leveled as 2) factors, A, B, C, D, and two suspected interactions, BC and CD. Thus to
determine the degrees of freedom or df, at least seven treatment conditions (experiments) are
needed for the two-level,
df=4(2-1)+2(2-1)(2-1)+1=7
Selecting the Orthogonal Array
Once the degrees of freedom are known, factor levels are identified, and possible interaction to
be studied, the next step is to select the orthogonal array (OA). The number of treatment
conditions is equal to the number of rows in the OA and must be equal to or greater than the
degrees of freedom. Table 3-5 shows the orthogonal arrays that are available, up to OA36. Thus,
if the number of degrees of freedom is 13, then the next available OA is OA16. The second
column of the table has the number of rows and is redundant with the designation in the first
column. The third column gives the maximum number of factors that can be used, and the last
four columns give the maximum number of columns available at each level.
Analysis of the table shows that there is a geometric progression for the two-level arrays of OA4,
OA8, OAI6, OA32, ... , which is 22, 23, 24, 25, ... . For the three-level arrays of OA9, OA27,
OA8I, ... , it is 32, 33, 34, ..... Orthogonal arrays can also be modified.
Table 3-5 Required Orthogonal Array
Interaction Table
Confounding is the inability to distinguish among the effects of one factor from another
factor and/or interaction. In order to prevent confounding, one must know which columns to use
for the factors in Taguchi method. This knowledge is provided by an interaction table, which is
shown in Table 3-6.
Table 3-6 Interaction Table for OA8
Let's assume that factor A is assigned to column 1 and factor B to column 2. If there is an
interaction between factors A and B, then column 3 is used for the interaction, AB. Another
factor, say, C, would need to be assigned to column 4. If there is an interaction between factor A
(column 1) and factor C (column 4), then interaction AC will occur in column 5. The columns
that are reserved for interactions are used so that calculations can be made to determine whether
there is a strong interaction. If there are no interactions, then all the columns can be used for
factors. The actual experiment is conducted using the columns designated for the factors, and
these columns are referred to as the design matrix. All the columns are referred to as the design
space.
Linear Graphs Taguchi developed a simpler method to work with interactions by using linear graphs.
Figure 3-17 Two linear Graphs for OA8
Two linear graph are shown in Figure 3-17 for OA8. They make it easier to assign factors and
interactions to the various columns of an array. Factors are assigned to the points. If there is an
interaction between two factors, then it is assigned to the line segment between the two points.
For example, using the linear graph on the left in the figure, if factor B is assigned to column 2
and factor C is assigned to column 4, and then interaction BC is assigned to column 6. If there is
no interaction, then column 6 can be used for a factor.
The linear graph on the right can be used when one factor has three two-level or higher order
interactions. Three-level orthogonal arrays must use two columns for interactions, because one
column is for the linear interaction and one column is for the quadratic interaction. The linear
graphs-and, for that matter, the interaction tables-are not designed for three or more factor
interactions, which are rare events. Linear graphs can also be modified. Use of the linear graphs
requires some trial-and-error activity
Interactions
Interactions simply means relationship existing between different X-factors/X with noise
variables considered for experiment. Figure 3-18 shows graphical relationship between any two
factors. At (a) there is no interaction as the lines are parallel; at (b) there is little interaction
existing between the factors; and at (c) there is a strong evidence of interaction. The graph is
constructed by plotting the points A1B1 A2B2, A2B1 and A2B2.
Figure 3-18 Interaction between Two Factors Signal-to-Noise (SIN) Ratio
The important contribution of Taguchi is proposing the signal-to-noise (S/N) ratio. It was
developed as a proactive equivalent to the reactive loss function. When a person puts his/her foot
on the brake pedal of a car, energy is transformed with the intent to slow the car, which is the
signal. However, some of the energy is wasted by squeal, pad wear, and heat. Figure 3-19
emphasizes that energy is neither created nor destroyed.
Figure 3-19 Concept of Signal-to-Noise (S/N) Ratio
Signal factors (Y) are set by the designer or operator to obtain the intended value
of the response variable. Noise factors (S2) are not controlled or are very expensive or difficult to
control. Both the average, y, and the variance, s2, need to be controlled with it single figure of
merit. In elementary form, S/N is /y s , which is the inverse of the coefficient of variation and a
unit less value. Squaring and taking the log transformation gives
( )2 210/ logNS N y s= −
Adjusting for small sample sizes and changing from Bels to decibels for NTB type gives
( ) ( )2 210/ 10log 1 /NS N y s n = − −
There are many different S/N ratios. The equation for nominal-the-best was given above. It is
used wherever there is a nominal or target value and a variation about that value, such as
dimensions, voltage, weight, and so forth. The target ( )τ is finite but not zero. For robust
(optimal) design, the S/N ratio should be maximized. The-nominal-the-best S/N value is a
maximum when the average is near target and the variance is small. Taguchi's two-step
optimization approach is to identify factors (X) which reduces variation of Y, and then bring the
average (Y) on target by a different set of factor (X). The he S/N ratio for a process that has a
temperature average of 21°C and a sample standard deviation of 2°C for four observations is
given by
( ) ( )
( ) ( )
2 210
2 210
/ 10log 1 /
=10log 21 2 1 / 4
=20.41 dB
NS N y s n = − − − −
The adjustment for the small sample size has little effect on
the answer. If it had not been used, the answer would have been 20.42 dB.
Smaller-the-Better
The S/Ns ratio for smaller-the-better is used for situations where the target value ( )τ is
zero, such as computer response time, automotive emissions, or corrosion. The S/N equation
used is
( )210 10/ 10log 10log /SS N MSD y n = − = − Σ
The negative sign ensures that the largest S/N value gives the optimum value for the
response variable and, thus a robust design. Mean square deviation (MSD) is given
to show the relationship with the loss function.
Larger-the-Better
The S/N ratio for larger-the-better type of characteristic is given by
( )210 10/ 10log 10log 1 / /LS N MSD y n = − = − Σ
Let us consider a battery life experiment. For the existing design, the lives of three AA batteries
are calculated as 20, 22, and 21 hours. A different design produces batteries life of 17, 21, and 25
hours. To understand which is a better design (E or D) and by how much, we can use the S/N
ratio calculation. As it is a larger-the-better (LTB) type of characteristic (Response), the
calculation are
( ) = − Σ + + +
210
10 2 2 2
/ 10log 1 / /
1 1 1 =-10log /320 22 25
=26.42 dB
ES N y n
( ) = − Σ + + +
210
10 2 2 2
/ 10log 1 / /
1 1 1 =-10log /317 21 25
=26.12 dB
DS N y n
26.42 26.12 0.3 db∆ = − =
The different design is 7% better than existing design. More data will be required to confirm the
result and so-called ‘Confirmatory trials’.
Although the metric signal-to-noise ratio have achieved good practical results, they are yet to be
accepted universally as a valid statistical measure. The controversy is on measures and shape of
loss function. However, Taguchi’s concept has resulted in a paradigm shift in the concept of
product quality and can optimize without any empirical regression modeling concept.
It is also to be noted that inner (controllable factors) and outer array (for noise variable) design is
recommended by Taguchi to understand the best setting for Robust Design, which many a times
researchers omit for ease of experimentation. This practice may be avoided. Engineering
knowledge and idea of interaction is essential to get the best benefit out of OA design. For further
details on Taguchi method, reader may refer the books written by P J Ross (1996), A Mitra
(2008),Besterfield et at. (2004) and M Phatke (1995).
Module III Product Quality Improvement
Lecture 5 - How six sigma philosophy is aligned with product quality improvement?
In 2000, M. Harry and R. Schroeder published ‘Six Sigma: The Breakthrough Management
Strategy Revolutionizing the World's Top Corporations’. Since that time, there has
been considerable interest in this subject. In this book, the authors devoted much space to a
review of the concept. In the Six Sigma world, the Quality Planning Process is referred to as
Design for Six Sigma (DFSS). DFSS is focused on creating new or modified product designs that
are capable of significantly higher levels of performance (using Six Sigma Methodology). They
emphasized on a Define-Measure-Analyze-Design-Verify (DMADV) sequence of quality
planning and design methodology that can be used for product or service desing. The DFSS
matrix is a tool which captures the important quality planning information that allows a six
sigma team to record the vital planning information and deliver as required in the DMADV
phases.
Statistical Concept for Six Sigma
According to James Harrington, "Six sigma is simply a TQM process that uses process capability
analysis as a way of measuring improvement”. Sigma, σ , is the Greek symbol for the statistical
measurement of dispersion, so-called standard deviation. It is the best measurement of process
variability, as smaller the deviation value, the less variability is there in the process. Figure 3.20
shows measurement (Y characteristic) on samples collected from that is normally distributed and
centered exactly on target, having the upper and lower specification limit (USL and LSL). The
estimated ± 6σ fits exactly with specification limit. For such situation, 99.9999998% of the
product or service will be between specifications, and the nonconformance rate will be 0.002
parts per million, or 2.0 parts per billion. The situation diagrammed represents a process
capability index (Cp & Cpk) of 2.0. A Cpk of 1.33 has been considered in industry as a de-facto
standard earlier. Table 3-7 shows the percent between specifications, the nonconformance rate,
and process capability for different specification limit locations.
Figure 3-20 Normal Distribution and Specification bound for a Quality Characteristics
Table: 3-7 Process centered on Target
Specification limit Percent conformance
Nonconformance
rate(ppm)
Process capability
(Cpk)
±1σ 68.70 317300 0.33
±2σ 95.45 485500 0.67
±3σ 99.73 2700 1
±4σ 99.9937 63 1.33
±5σ 99.999943 0.57 1.67
±6σ 99.9999998 0.002 2
According to the six-sigma philosophy, any process rarely stay centered-the center tends to
"shift" above and below the target, µ . Figure 3-21 shows a process that is normally distributed,
but has shifted within a range of 1.5σ above and 1.5σ below the target. For the Figure 3-21
situation, 99.9996600% of the product output or service output characteristic will be between
specifications and the nonconformance rate will be 3.4 ppm. The off-center situation gives a
process capability index (Cpk) of 1.5 with 1.33 being the defacto standard previously. Table 3-8
shows the percent between specifications, the nonconformance rate, and capability for different
specification limit locations for an off-centered process. The magnitude and type of shift is a
matter of analysis and should not be assumed ahead of time. There is rare evidence of case
studies in literature that indicates a shift more than 1.50 σ. The automotive industry recognized
the concept of Six Sigma in the mid-1980's, evaluated it and deemed it acceptable. It is to be
noted that the original work on six sigma was based on a few empirical studies of a single
output.
Figure 3-21 Shift in process output characteristic mean
Table: 3-8 Process off-centered by 1.5 σ
Specification limit
Percent
conformance
Nonconformance
rate(ppm)
Process capability
(Cpk)
±1σ 30.23 697700 -0.167
±2σ 69.13 308700 0.167
±3σ 93.32 66810 0.5
±4σ 99.379 6210 0.834
±5σ 99.9767 2330 1.167
±6σ 99.9966 3.4 1.5
The statistical aspects of six-sigma tell us that we should reduce the process variability, σ, and try
to keep the process centered on the target,µ . These concepts are not new, and had been long
advocated by Shewhart, E Deming, and G Taguchi.
Six Sigma DMAIC Methodology
The standard problem-solving approach used in Six Sigma is known as DMAIC(Define,
Measure, Analyze, Improve, and Control).
Define Phase
After a Six Sigma project is selected, the first step is to clearly define the problem. This
activity is significantly different from project selection. Project selection generally
responds to symptoms of a problem and usually results in a rather abstract problem
statement. One must describe the problem in operational or measurable terms that facilitate
further analysis. For example, a firm might have a history of poor reliability of electric
generator it manufactures, resulting in a Six Sigma project to improve generator reliability.
A preliminary investigation of warranty and field service repair data might suggest that the
source of most problems are brush wear, and more specifically, suggest a problem with
brush hardness variability. Thus, the problem might be defined as "reduce the variability of
brush hardness." This process of drilling down to a more specific problem statement is
sometimes called project scoping.
A good problem statement also should identify the customer (external or internal) and the
CTQ (Critical-to-Quality) Characteristics that have the most impact on product or service
performance, describe the current level of performance or the nature of errors or customer
complaints, identify the relevant performance metrics, benchmark best performance
standards, calculate the cost/ revenue implications of the project, and quantify the expected
level of performance from a successful Six Sigma effort. The Define phase should also
address such project management issues as what will need to be done, by whom, and
when.
Measure
This phase of the DMAIC process focuses on how to measure the internal
processes that impact CTQ’s. It requires an understanding of the causal relationship(s)
between process performance and customer concept of value. However, once
they are understood, procedures for gathering facts-collecting reliable data or observations,
and careful listening-must be defined and implemented. Data from existing production
processes and practices often provide important information, as does feedback from
supervisors, workers, customers, and field service employees. An important concept
required at this stage is ‘Opportunity’ and ‘Rolled Throughput Yeild’
(http://asq.org/qic/display-item/?item=15398).
Analyze
A major flaw in many problem-solving approaches is a lack of emphasis on rigorous
statistical analysis. Too often, we want to jump to a solution without fully understanding
the nature of the problem and identifying the source of the problem. The Analyze phase of
DMAIC focuses on why defects, errors, or excessive variation occur.
After potential causal variables are identified, statistical experiments are conducted to
verify them. These experiments generally consist of formulating some hypothesis to
investigate, collecting data, analyzing the data, and reaching a reasonable and statistically
supportable conclusion. Statistical thinking and analysis plays a critical role in this phase.
It is one of the reasons why statistics plays an important part in Six Sigma training.
Improve
Once the root cause of a problem is understood, the analyst or team needs to generate
ideas for removing or resolving the problem and improve the performance measures or
CTQ. This idea-gathering phase is a highly creative activity, because many optimal
solutions are not obvious. One of the difficulties in this task is the natural instinct to
prejudge ideas before thoroughly evaluating them. Most people have a natural fear of
proposing a "silly" idea or looking foolish. However, such ideas may actually form the
basis for a creative and useful solution. Effective problem solver must learn to defer
judgment and develop the ability to generate a large number of ideas at this stage of the
process, whether practical or not.
After a set of ideas have been proposed, it is necessary to evaluate them and select the
most promising. This process includes confirming that the proposed solution will
positively impact the key process output variables or CTQ, and identifying the maximum
acceptable ranges of these variables.
Problem solutions often entail technical or organizational changes. Often some sort of
decision model is used to assess possible solutions against important criteria such as cost,
time, quality improvement potential, resources required, effects on supervisors and
workers, and barriers to implementation such as resistance to change or organizational
culture. To implement a solution effectively, responsibility must be assigned to a person or
a group who will follow through on what must be done, where it will be done, when it will
be done, and how it will be done.
Control
The Control phase focuses on how to maintain the improvements, which includes putting
statistical process control (SPC) in place to ensure that the key variables remain within
the naturally acceptable limits under the modified process. These improvements might
include establishing the new standards and procedures, training the workforce, and
instituting controls to make sure that improvements do not die over time. Controls might
be as simple as using checklists or periodic status reviews to ensure that proper
procedures are followed.
Overall, Six Sigma Methodology is to work smarter not harder. It also emphasizes on
measurement(s) that impact customer, ways to improve the process, and decision making
based on firm statistical concept. Reader may refer any standard book/references given
below to learn further on Six Sigma Methodology.